AXIS 2.

4 USER GUIDE
SOFTWARE MANUAL FOR THE
MAESTRO™ AND MUSE™
MEA SYSTEMS
Trademarks

Axion BioSystems, Inc. and are trademarks of Axion BioSystems, Inc., and may not be
used without the express written permission of Axion Biosystems, Inc.

All other brands, product names, company names, trademarks, and service marks are the properties of
their respective owners.

Restrictions and Liabilities

This document is provided “as is” and Axion BioSystems will assume no responsibility for any
typographical, technical or other inaccuracies in this document. Axion BioSystems does not make any
commitment to provide any changes, updates, or enhancements to this document within any time frame or
at all.

This document might contain references to third-party sources of information, hardware or software,
products or services and/or third-party websites (collectively the “Third-Party Information”). Axion
BioSystems has no control over and is not responsible for any Third-Party Information, including without
limitation the content, ac- curacy, copyright compliance, compatibility, performance, trustworthiness,
legality, decency, links or any other aspect of Third-Party Information. The inclusion of Third-Party
Information in this document does not imply endorsement with the use of Axion BioSystems technology.

Conditions of Use

The user is responsible for understanding and performing the protocols that are described within. Axion
BioSystems makes no guarantee for any results. These protocols are provided as a recommendation by
Axion BioSystems based on use and experience.

Copyright Notice

Copyright © 2017 Axion BioSystems, Inc. All rights reserved. This document may not be reproduced,
distributed, modified or publicly displayed without the express written permission of Axion BioSystems.
TABLE OF CONTENTS
Chapter 1. Introduction............................................................................................................................................... 1
1.1. New AxIS Features ......................................................................................................................................... 1
1.2. Standalone Tool Updates .............................................................................................................................. 2
1.3. Technical Support ........................................................................................................................................... 3
Chapter 2. AxIS Overview ......................................................................................................................................... 4
2.1. Menu Bar ......................................................................................................................................................... 4
2.2. Active Plate ...................................................................................................................................................... 5
2.2.1. Plate Map Editor ..................................................................................................................................... 5
2.2.2. Enabling or Disabling Wells and Electrodes ......................................................................................... 6
2.2.3. Entering Well Information ....................................................................................................................... 6
2.2.4. Setting Well Color ................................................................................................................................... 7
2.2.5. Importing and Exporting Plate Maps ..................................................................................................... 8
2.2.6. Creating a Plate Map in Excel ............................................................................................................... 9
2.3. Streams .......................................................................................................................................................... 10
2.3.1. Data Processors ..................................................................................................................................... 11
2.3.2. Digital Filter ............................................................................................................................................ 12
2.3.3. Artifact Eliminator .................................................................................................................................. 12
2.3.4. Stimulation Inspector ............................................................................................................................. 12
2.3.5. Spike Detector ....................................................................................................................................... 13
2.3.6. Burst Detector ........................................................................................................................................ 15
2.3.7. Cardiac Beat Detector .......................................................................................................................... 16
2.3.8. Cardiac Statistics Compiler .................................................................................................................. 19
2.3.9. Neural Statistics Compiler .................................................................................................................... 21
2.4. File Play and Display Controls ..................................................................................................................... 22
2.5. Status Bar ....................................................................................................................................................... 22
2.5.1. Adding Timestamp Notes ..................................................................................................................... 23
2.6. Control Bar .................................................................................................................................................... 24
2.6.1. Experiment Setup Properties ................................................................................................................. 24
2.6.2. Scheduled Recording Setup ................................................................................................................. 25
2.6.3. Stimulation Studio .................................................................................................................................. 27
 2.6.4. Environmental Control........................................................................................................................... 28
2.6.5. Statistics Graphing ................................................................................................................................ 29
2.6.6. Activity Map .......................................................................................................................................... 30
2.6.7. Continuous Waveform Plots ................................................................................................................. 31
2.6.8. Spike Plots .............................................................................................................................................. 32
2.6.9. Cardiac Beat Plots ................................................................................................................................. 34
Chapter 3. Data Acquisition ..................................................................................................................................... 37
3.1. AxIS Communication with the Maestro/Muse ........................................................................................... 37
3.2. Configuring the Maestro/Muse for Acquisition ......................................................................................... 38
3.2.1. Configuring the Hardware ................................................................................................................... 38
3.2.2. Stream Configurations .......................................................................................................................... 39
3.3. Live Data Acquisition Tutorial ....................................................................................................................... 41
3.3.1. Setting Up for a Recording ................................................................................................................... 41
3.3.2. Recording from the Maestro Manually ............................................................................................... 42
3.3.3. Recording from the Maestro using the Scheduled Recorder ............................................................. 43
3.3.4. Recording from the Muse ..................................................................................................................... 43
Chapter 4. Stimulation .............................................................................................................................................. 44
4.1. Electrical Stimulation ..................................................................................................................................... 45
4.1.1. Electrical Stimulation Blocks ................................................................................................................. 46
4.1.2. Stimulation Protocol Generation .......................................................................................................... 47
4.1.3. Example Neural Stimulation Patterns................................................................................................... 49
4.1.4. Cardiac Pacing Stimulus Design Workflow......................................................................................... 50
4.2. Optical Stimulation ....................................................................................................................................... 52
4.2.1. Optical Stimulation Blocks .................................................................................................................... 53
4.2.2. Stimulation Protocol Generation .......................................................................................................... 53
4.2.3. Example Stimulation Pattern ................................................................................................................. 56
4.3. Evaluating Cardiac Stimulation Capture ..................................................................................................... 56
4.4. Evaluating Neural Stimulation Capture....................................................................................................... 59
Chapter 5. Data Analysis ......................................................................................................................................... 60
5.1. Stream Configurations .................................................................................................................................. 60
5.2. Output File Types .......................................................................................................................................... 61
5.2.1. Recorded File Names ......................................................................................................................... 63
 5.3. Preparing Files for Analysis .......................................................................................................................... 65
5.4. Selecting the Analysis Window ................................................................................................................... 65
5.4.1. Re-recording Segments of .raw Files ................................................................................................... 66
5.5. Data Analysis Tutorial ................................................................................................................................... 67
5.6. Analysis of Multiple Files (Batch Processing) .............................................................................................. 68
Chapter 6. Cardiac Data Analysis........................................................................................................................... 70
6.1. Cardiac Activity Introduction........................................................................................................................ 70
6.2. Identifying Cardiac Depolarization and Repolarization ........................................................................... 72
6.2.1. Cardiac Beat Detection ........................................................................................................................ 72
6.2.2. T-Wave Identification............................................................................................................................ 72
6.3. Cardiac Statistics Compiler Endpoints ........................................................................................................ 73
Chapter 7. Neural Data Analysis ............................................................................................................................ 75
7.1. Neural Activity Introduction ......................................................................................................................... 75
7.2. Identifying Neural Spikes ............................................................................................................................. 75
7.3. Types of Neuronal Activity ........................................................................................................................... 77
7.4. Neural Statistics Compiler Endpoints .......................................................................................................... 79
Appendix A. Axion Standalone Tools ..................................................................................................................... 81
Appendix B. Neural Metric Tool .............................................................................................................................. 82
B.1. Introduction.................................................................................................................................................... 82
B.2. Neural Metric Tool Overview ...................................................................................................................... 82
B.2.1. Plate Map Display and Analysis Parameters ...................................................................................... 82
B.2.2. Bursting Display and Parameters.......................................................................................................... 84
B.2.3. Synchrony Display and Parameters ..................................................................................................... 86
B.2.4. Evoked Display and Parameters .......................................................................................................... 87
B.2.5. Advanced Options ................................................................................................................................ 88
B.3. Operation ...................................................................................................................................................... 89
B.3.1. Generate AxIS Spike file ...................................................................................................................... 89
B.3.2. Analyze a single AxIS Spike file .......................................................................................................... 90
B.3.3. Analyze multiple AxIS Spike files ......................................................................................................... 90
B.4. Output ............................................................................................................................................................ 91
Appendix C. CiPA Analysis Tool .............................................................................................................................. 94
C.1. Introduction ................................................................................................................................................... 94
 C.2. CiPA Analysis Tool Overview ...................................................................................................................... 94
C.2.1. Golden Channel Selector..................................................................................................................... 95
C.2.2. FPD Detection Display .......................................................................................................................... 96
C.2.3. Arrhythmia Inspector............................................................................................................................. 97
C.3. Operation ...................................................................................................................................................... 99
C.3.1. Process Recordings in AxIS .................................................................................................................. 99
C.3.2. Analyze Data in the CiPA Analysis Tool .......................................................................................... 100
C.3.3. To load a saved experiment ............................................................................................................. 101
C.4. Outputs ....................................................................................................................................................... 102
C.4.1. Export Figures..................................................................................................................................... 102
C.4.2. Export Plot Source Data to CSV........................................................................................................ 103
C.4.3. Export Well Endpoints to CSV .......................................................................................................... 103
C.4.4. Export CiPA CSV ................................................................................................................................ 103
Appendix D. AxIS Metric Plotting Tool ................................................................................................................. 104
D.1. Introduction ................................................................................................................................................ 104
D.2. Operation ................................................................................................................................................... 104
D.2.1. Load files for analysis ........................................................................................................................ 104
D.2.2. Adjust data display settings ............................................................................................................... 105
D.2.3. Inspect plots ........................................................................................................................................ 106
D.2.4. Export Figures and Files ..................................................................................................................... 107
D.3. Output......................................................................................................................................................... 107
Appendix E. Axion Data Export Tool .................................................................................................................... 109
E.1. Introduction................................................................................................................................................ 109
E.2. Operation .................................................................................................................................................. 109
E.3. Output ......................................................................................................................................................... 109
 Introduction

CHAPTER 1. INTRODUCTION
Axion Integrated Studio (AxIS) is a multipurpose software for data acquisition and analysis with the Maestro
and Muse microelectrode array (MEA) systems. This manual provides a basic overview of AxIS and
instructions for data acquisition, stimulation, and analysis. For step-by-step instructions for data acquisition and
analysis, see Sections 3.3 and 5.5. For operation of the Maestro, Muse, Lumos, APEX, or ECMini, refer to
their respective manuals. For cell culturing protocols, application notes, posters, and publications, visit
www.axionbiosystems.com. Sample data files can be found at Libraries/Documents/Public
Documents/Axion BioSystems/Manuals and Documentation/Sample Data on the Maestro/Muse computer.

Additional analysis tools are available to supplement AxIS and provide application-specific analysis support
(see Appendix A). Tools are available for download through the ShareFile system. Tutorial videos covering
new AxIS and standalone tool features are also available through the ShareFile system in the Documentation
folder. Contact [email protected] for ShareFile access.

1.1. NEW AXIS FEATURES

AxIS 2.4 contains a variety of new features designed to improve electrical stimulation and support new plate
types, specifically CytoView and E-Stim+ MEA plates. The following new features are available in this release:

Electrical Stimulation:

 Dedicated Neural Stimulation Block: Built-in optimized neural stimulation parameters and improved
artifact management allow easier identification of evoked neural responses (Section 4.1).
 Electrical Stimulation Tags: The timing of electrical stimulation events is stored to the raw file as tags,
enabling automated analysis of evoked activity (Section 4.1.1 and B.2.4).
 Cardiac Pacing using E-Stim+ MEA Plate: The cardiac pacing block now contains optimized stimulation
parameters and artifact management for electrical stimulation using the dedicated stimulation electrode on
the E-Stim+ plate, in addition to microelectrode stimulation (Section 4.1.1).
 Stimulation Inspector: A new visualization data processor provides a close look at data near the
stimulation tag times, to evaluate capture and ensure artifact elimination (Section 2.3.4).

Neural:

 Coincident event removal: The spike detector contains a coincident event removal option that can be used
to eliminate non-physiological plate-wide or well-wide artifacts (Section 2.3.5).
 Raster plots associated with spike detector: The raster will be plotted in the Spike Plots view regardless of
whether or not a Burst Detector is added to the stream (Section 2.6.8).
 Updated metrics included in Treatment Averages: Treatment Averages in the Neural Statistics Compiler
csv file now include the Number of Active Electrodes and Weighted Mean Firing Rate (Section 7.4).

[1]
AxIS 2.4 User Guide

Usability:

 AxIS 2.4 fully supports CytoView and E-Stim+ MEA plates (Section 2.2)
 Live recordings and analysis files can use macros to automatically add descriptive information to file
names (Section 5.2.1)
 Stack Continuous Waveform Plots and Cardiac Beat Plots (Section 2.6.7 and 2.6.9)
 Import/export plate maps to and from Excel in color (Section 2.2.6)
 Dedicated configurations for spontaneous, electrically evoked, and optically evoked activity (Sections
3.2.2 and 5.1)
 Usability improvements in Stimulation Studio (Section 4.2.2):
 If there is only one lane of stimulation, clicking the “X” to delete the lane will clear the lane
 Copy/paste lanes in Stimulation Studio
 Hide a lane from the Compiled Pattern View at the top of the display
 APEX Remote Control improvements
 New Help option to Create Support Bundle, a zip file of log files and configurations to easily email the
Axion support team, enabling optimal and efficient support
 AxIS version number displayed in the title bar
 Double-click a raw file in Windows Explorer to open the file in AxIS

1.2. STANDALONE TOOL UPDATES

Axion BioSystems provides a variety of standalone tools for processing data and producing figures. Guides
for these tools can be found in the Appendix of this User Guide. The following list highlights changes to the
standalone tools with this release, as well as introduces a new standalone analysis tool – the CiPA Analysis
Tool.

Neural Metric Tool:

 Batch Processing
 Evoked activity analysis for electrical stimulation
 Envelope Network Burst Detection, better suited for tonically active cultures
 Improvement to ISI Adaptive algorithm for Network Burst Detection
 Zoom controls in raster
 Tag-based blanking for electrical and optical stimulation artifact management
 Plate map included in output csv files
 Advanced option to exclude network bursts in specified wells
 New metrics: Network ISI CoV and ISI CoV within Network Bursts
 Improved file segment options under Analysis Parameters to mirror AxIS

AxIS Metric Plotting Tool:

 Export plot source data in the same layout as displayed on the plots
 New machine-readable output file for easier import into statistical analysis programs

[2]
 Introduction

CiPA Analysis Tool:

 Quick and easy comparison of beating and waveforms between recordings and doses in a user-friendly,
visual display
 Inspection and verification of field potential measurements
 Easy detection and classification of arrhythmia
 Endpoint reporting in a variety of formats, including the CiPA format as specified by the FDA
 Summary figures and reports to compile experiment results

Note: The ‘Conduction Velocity Tool’, the ‘Cardiac Data Aggregation Tool’, and the ‘Cardiac Data Plotting
Tool’ will not be supported after ‘AxIS 2.4’. Please see Appendix A for more information, including where to
find the functionality of these tools in ‘AxIS’ and other standalone tools.

1.3. TECHNICAL SUPPORT

For additional AxIS support and assistance, please contact your authorized Axion Biosystems distributor or
Axion Biosystems directly.

Axion Biosystems contact information:

Phone: 1 (404) 477-2557
Email: [email protected]

Please have the following information available when requesting technical assistance:

 Description of the problem

 What was happening at the time of the error
 What you have tried so far to solve the problem
 A copy of Axion Support Bundle (generate from the Help menu)
 Screen shots of any errors

[3]
AxIS 2.4 User Guide

CHAPTER 2. AXIS OVERVIEW

The AxIS user interface is designed to provide easy access to all experimental controls. The top Menu bar
controls file loading, saving, and display. Most of the settings for data acquisition, analysis and display are
found in the left panel. The left panel contains the current plate information (Active Plate), the File Play and
Display controls, and the data stream and data processors under Streams. The Control Bar along the bottom
contains functional and data visualization modules. The active window displays the module selected from the
Control Bar below. The Status Bar at the bottom of the active window displays file status, play and record
times, and timestamp notes.

Menu Bar

Active Plate

Display controls

Active window
File Play controls

Streams

Status Bar
Control Bar

2.1. MENU BAR

The Menu bar has four options:

Menu Option Description

File Open Recording(s)… Loads one or more .raw files into the Streams window.
New Batch Process… Opens the batch processing dialog. See Section 5.6.
Loads a stimulation waveform file (.sswf) into the Stimulation
Open Stim Waveform
Studio module. See Chapter 4.
Saves the current stimulation waveform from Stimulation
Save Stim Waveform
Studio. See Chapter 4.
Exit Closes AxIS.
View Show All Modules Displays all modules on the Control Bar. See Section 2.6.
Module Name Displays the selected Control Bar module in the active
window.
Tools Enable Remote Control Allows AxIS to receive commands from third party software.

[4]
 AxIS Overview

Help User Guide Opens the AxIS user guide.

Create Support Bundle Creates an error log report for email to Axion Support team.
About Displays AxIS version number and information.

2.2. ACTIVE PLATE

The Active Plate displays the current plate configuration. The plate Select Plate Type
configuration determines how electrodes are mapped to the wells
and is critical for proper analysis. The Active Plate interface can be
used to select wells or electrodes for recording and store
information about well contents. Once a plate type is selected, that
plate type is retained in AxIS until a new one is selected.

Use the Active Plate drop-down menu to select the correct plate
type prior to recording. If a file is recorded with the incorrect plate Active Plate
type, it may be changed in the output file.

To change the plate type on a previously recorded file:

1. Click File  Open Recording(s).

2. Select the desired file.
3. Click the Lock icon to the right of the Active Plate drop-down menu.
4. Select the desired plate type from the Active Plate drop-down menu
Note: If the plate type is changed to a plate type of a different format, the plate map information will be
cleared.
The Active Plate drop-down menu has the following options:

Option Plate name Part number

Classic MEA 48 (incl. Classic MEA 48 M768-KAP-48
AccuSpot) AccuSpot MEA 48 M768-KAP-48A
E-Stim + Classic MEA 48 E-Stim+ Classic MEA 48 M768-KAP-48S
Classic MEA 96 Classic MEA 96 M768-KAP-96
CytoView MEA 12 M768-GL1-30Pt200
CytoView MEA 12
M768-GL1-30Au200
CytoView MEA 48 M768-tMEA-48B
CytoView MEA 48
M768-tMEA-48W
Lumos MEA 48 Lumos MEA 48 M768-tMEA-48OPT

2.2.1. Plate Map Editor

Double-click the Active Plate to open the Plate Map Editor. The Plate Map Editor is used to select wells or
electrodes for recording and store information about well contents. The editor has a representation of the plate

[5]
AxIS 2.4 User Guide

where each well consists of a square of circle electrodes. The right contains well information of the currently
selected well(s), including the Treatment, Concentration, Well Color, Notes, and whether the well is on or off
for recording. Import and Export buttons allow the user to load previously created plate maps or export the
current map for future use. The Lock icon prevents the information from being edited. When editing the plate
map associated with the live Maestro stream, the Plate Map Editor is always unlocked. For a previously
recorded raw file, click the lock on the lower left to make changes. Information entered in the editor is stored
with the recording and used in data analysis.

Note: All changes to the plate map are automatically saved to a previously recorded file.

Well On/Off Status Well Color

Selected Well

Treatment Information

Electrode Off

Unlock to Edit

2.2.2. Enabling or Disabling Wells and Electrodes

Once selected, wells can be turned on or off with the toggle switch under Well Information. Disabled wells
appear black in the Plate Map Editor. An electrode may be turned on or off by double-clicking on it. Disabled
electrodes are connected to ground and appear gray in the Plate Map Editor. Turning off a well or electrode
will prevent it from being recorded or analyzed, reducing the file size.

2.2.3. Entering Well Information

Click on a well to select it. Multiple wells can be selected by clicking on the column or row label to highlight
the column or row, respectively; clicking the * in the upper left corner to highlight the entire plate; holding the
Ctrl key and clicking on the desired wells; selecting two wells while holding the Shift key to select all of the
wells between them; or by click-and-drag selecting wells in a region.

[6]
 AxIS Overview

Note: A plate map can only be edited when it is unlocked using the Lock button.
To assign well information:

1. Select the desired well(s).

2. Type a name in the Treatment field and press enter or click on the green check mark that appears. Do
not use commas in the Treatment name. Use the control checkbox to indicate if this well is a control
well.
3. Type the concentration in the Concentration field, including units (a space between the units and
concentration is optional), and press enter or click on the green check mark that appears. Do not use
commas, dashes, or special characters in the Concentration field. If concentration is not applicable,
leave the field blank.
Note: ‘AxIS’ will convert “uM” or “um” to “µM” or “µm”, respectively.
4. Enter any additional information in the Notes field.
Note: When selecting multiple wells to edit, each field operates independently. If only the Treatment is
entered, Concentration information will remain unchanged for those wells, and vice versa. If a field is
empty, it will not be cleared, rather remain unchanged.

To copy well information between wells:

1. Select the desired well(s).

2. Press Ctrl-C or right-click  Copy Well.
3. Select the destination well(s)
4. Press Ctrl-V or right-click  Paste Well
Note: Multiple wells can be copied and pasted at the same time, as long as the selected wells have the
same shape (e.g. copy/pasting a full row to a different row).

To clear well information from wells:

1. Select the desired well(s).

2. Click Clear Well.

2.2.4. Setting Well Color

To set well color manually:

1. Select the well(s).

2. Click on Well Color in the upper right corner of Well Information and choose a color.

To set well color automatically:

1. Right-click on any well.

2. Select Auto Color. Each treatment will receive a different hue and each concentration of a treatment
will receive a different shade.

[7]
AxIS 2.4 User Guide

To change the hue of a treatment group:

1. Right-click on any well.

2. Select the treatment group from the menu.
3. Select Set Hue and choose the hue from the options available. Each concentration in that treatment
group will be receive a different shade of the selected hue.

Right-click Menu

Clear Well
Import/Export Plate Map

2.2.5. Importing and Exporting Plate Maps

Plate maps may be imported or exported using the Import and Export buttons in the Plate Map Editor. Plate
maps may be copied and pasted to and from Microsoft Excel (Section 2.2.6).

To export a plate map using the export button:

1. Click the Export button in the Plate Map Editor.

2. Enter a name for the plate map and click Save. The file will be saved with a .platemap extension.

To import a plate map using the import button:

1. Click the Import button in the Plate Map Editor.

2. Select the desired .platemap, .raw, or .spk file and click Open.
Note: Selecting a previous recording (.raw file) or Spike file (.spk) will load its plate map into the
current file.

[8]
 AxIS Overview

2.2.6. Creating a Plate Map in Excel

AxIS recognizes two text formats which may be imported from or exported to Microsoft Excel.

In format 1, each row corresponds to a different piece of plate map data and each column represents a well.
The columns are labeled with the well ID where well A1 is row A, column 1. The plate map data includes:

Option Description
Active Whether or not the well is enabled. (TRUE or FALSE)
Well Coloring Color code for the Well Color
Control Whether or not Control field is selected (TRUE or FALSE)
Treatment String containing the Treatment field
Concentration String containing the Concentration field
Additional Information String containing the Notes field.

To copy the plate map in AxIS and paste to Excel in format 1:

1. Right-click on one or more wells in the Plate Map Editor.

2. Select Copy Wells.
3. Click on any cell in Excel and press Ctrl+V.

To copy the plate map in Excel and paste in AxIS:

1. Highlight the desired wells in Excel, including the Well Number.

2. Press Ctrl+C.
3. Select the corresponding wells in the Plate Map Editor.
4. Right-click on one of the wells and select Paste to Well(s).

In format 2, each cell represents a well on the plate and contains information in the format of “Treatment
[Concentration]”. The color of the cell corresponds to the Well Color. Rows are labeled A, B, C… and
columns are labeled 1, 2, 3….

To copy the plate map in AxIS and paste to Excel in format 2:

1. Right-click on any well in the Plate Map Editor.

2. Select Copy Plate Mapping.

[9]
AxIS 2.4 User Guide

3. Click on any cell in Excel and press Ctrl+V.

To copy the plate map in Excel and paste in AxIS:

1. Highlight the entire plate map in Excel, including the row and column headers.
2. Press Ctrl+C.
3. Select all wells in the Plate Map Editor.
4. Right-click on one of the wells and select Paste to Well(s).

2.3. STREAMS

The Streams window displays all currently loaded data streams. A data stream contains the live or recorded
continuous voltage data from a Maestro or Muse and all associated data processors. See Section 2.3.1 for
more information on data processors. A stream is organized in a hierarchy beginning with the continuous
voltage data and then passing through the data processors in sequence. Combining processors into different
configurations provides customizable stream flows for different applications. Axion provides preset processor
configurations for specific cardiac and neural applications (Sections 3.2.2 and 5.1). Customized
configurations can be manually constructed by placing individual data processors on the stream (Section
2.3.1) and saved or loaded by right-clicking on the stream.

The first stream represents the Maestro or Muse continuous voltage data. When the Maestro or Muse is
connected, the stream will be called Maestro or Muse, respectively. If no Maestro or Muse is connected, the
stream will be inactive and say “No Device Connected”.

Only active streams may be viewed or analyzed in AxIS and only one stream may be active at a time. The
active stream is indicated by a colored square and bold black text. Inactive streams have gray squares and
gray text.

To activate a data stream:

1. Right-click on the continuous voltage data name.

2. Select Select for Play/Rec.

To load a new data stream:

1. Click File Open Recording(s)….

2. Navigate to the desired recording (.raw file) and click Open.

To remove a data stream or batch process:

1. Right-click on the continuous voltage data name.

2. Select Remove.

To remove all data streams in the Streams window:

1. Right-click on white space in the Streams window.

[10]
 AxIS Overview

2. Select Remove All Files.

2.3.1. Data Processors

Data processors are added to data streams to alter or analyze the data. They all require input from a
continuous data stream or another data processor, and output an altered continuous data stream, metric(s),
and/or an output file. Information from some data processors are visualized using the various AxIS modules.
Some data processors and settings are disabled during live data streams and some may not be modified
unless streaming has stopped.

Preset configurations containing streams of data processors recommended for most applications are provided
in AxIS. See Sections 3.2.2 and 5.1 for more information about recommended default stream configurations.

To add a data processor:

1. Right-click on the stream in the Streams window.

2. Select Add Processing and select the desired data processor.
Note: Some data processors are dependent on others and may only be added below the requisite
processor.
3. Enter the desired settings and click OK.

To remove a data processor:

1. Right-click on the processor’s stream in the Streams window.

2. Select Remove.
Note: Removing a data processor will remove all processors below it on the same continuous data
stream.

To access the settings of a data processor:

1. Right-click on the processor’s stream in the Streams window.

2. Select Settings.
-- Or --
1. Double-click on the processor’s stream in the Streams window.

To change the visibility of a data processor:

1. Right-click on the processor’s stream in the Streams window.

2. Select Plot  Show. If enabled, the processor will show in dependent AxIS modules.

To change the color of a data processor:

1. Right-click on the processor’s stream in the Streams window.

2. Select Plot  Change Trace Color.
3. Select the desired color and click Enter.

[11]
AxIS 2.4 User Guide

2.3.2. Digital Filter

The Digital Filter processor applies a digital Butterworth bandpass filter to any continuous data stream to
further reduce noise. This filter is applied on top of hardware-based filters selected in the Maestro/Muse
Settings (Section 3.2.1) during a recording. The output of this data processor is a filtered continuous data
stream (.raw file), and data processors placed below the Digital Filter will act on the filtered data stream.

The digital filter has a High Pass Filter, the frequency components of the signal must be above to pass, and a
Low Pass Filter, the frequency components of the signal must be below to pass. To adjust either, enter the new
value into the respective Cutoff Frequency text field. By default the frequency range for neural configurations is
200 Hz (High Pass Filter) to 3 kHz (Low Pass Filter). For cardiac configurations the range is 0.1 Hz to 2 kHz.

2.3.3. Artifact Eliminator

The Artifact Eliminator removes artifact in raw signals resulting from electrical stimulation. The removal includes
blanking of the large initial impulse as well as removal of residual artifact based on the similarity of its profile
across electrodes. The output of this data processor is a continuous data stream (.raw file) with the artifact
removed, and data processors placed below the Artifact Eliminator will act on that output.

2.3.4. Stimulation Inspector

The Stimulation Inspector provides a close look at data near an Electrical Stimulation Tag to evaluate artifact
elimination and cell response to the stimulation. The Stimulation Inspector plots the continuous voltage data just
before and after an Electrical Stimulation Tag in the Spike Plots module. Note the Stimulation Inspector is for
visualization only and does not produce an output file.

[12]
 AxIS Overview

While using the Stimulation Inspector data processor, the Electrical Stimulation Tag will appear as a triangle
and vertical line at time zero on the spike plot. The time displayed before and after the tag may be changed in
the Stimulation Inspector settings by the Pre-Tag and Post-Tag fields, respectively. The waveform plots overlay
with the brightest trace representing the most recently detected stimulation event.

2.3.5. Spike Detector

The Spike Detector detects threshold crossings in the continuous data stream. These crossings are referred to
as “spikes” and are plotted in the Spike Plots module as spike waveform and raster plots. The Spike Detector
processor is the base processor for neural data visualization and analysis. It identifies the spike timing and
location and is the source for the Spike Plots, Activity Map, and Statistics Graphing modules. Burst Detector
data processors (Section 2.3.6) may only be placed below a Spike Detector. The Spike Detector produces
several outputs including AxIS Spike (.spk), Spike Count (.csv), and Spike List (.csv) files. See Section 5.2 for
more information about these output formats.

[13]
AxIS 2.4 User Guide

Setting a proper detection threshold is crucial for accurate neural data analysis. Lower thresholds increase the
incidence of false-positives (small noise events misidentified as spikes); higher thresholds may not detect
smaller amplitude action potentials. Axion recommends an adaptive threshold of 6 x standard deviations to
minimize both false-positives and missed detections.

The Spike Detector has four possible threshold detection methods selected by the Methods drop-down menu:

1. Adaptive Threshold (Recommended): Threshold is set on a per electrode basis, as a multiple of the
noise of the continuous data, each electrode threshold is specific to that electrode. The standard
deviation multiple can be set in the Threshold field. Detect Only Crossings requires the signal to return
below threshold before detecting an additional spike. The spike time is marked at the maximum slope
of the spike voltage waveform.
2. Static Threshold: Threshold is set on a plate-wide basis, all electrode thresholds are the same. Spikes
are detected as any event greater than the value defined in the Threshold field. Detect Only Crossings
requires the signal to return below threshold before detecting an additional spike. The spike time is
marked at the maximum slope of the spike voltage waveform.
3. Peak Detection Adaptive Threshold: Functions the same as the Adaptive Threshold above, but the
spike time is marked at the peak of the spike voltage waveform as defined by the Peak Detection
drop-down menu.
4. Peak Detection Static Threshold: Function the same as the Static Threshold above, but the spike time is
marked at the peak of the spike voltage waveform as defined by the Peak Detection drop-down
menu.
The Peak Detection drop-down menu options are:

Option Description
Positive Inflection Marks the first peak with a positive voltage
Negative Inflection Marks the first peak with a negative voltage
First Peak Marks the first peak regardless of polarity (positive or negative)
Maximum Amplitude Marks the peak with the largest amplitude

The Durations dialog box sets the display in the Spike Plots module. Use the Pre-Spike and Post-Spike fields to
specify how much time before and after each spike crossing will be displayed on the plot and saved as spike
waveforms (AxIS Spike files). The spike time, determined by the Method drop-down menu is plotted as zero.

The Coincident Events dialog box removes coincident artifacts from analysis results. Coincident artifacts are
artificial spikes detected on multiple electrodes at exactly the same time. They may occur during an electrical
stimulus or due to environmental interference like bumping the system or touching the media in a well. AxIS
identifies a coincident artifact as spikes occurring on a minimum number of electrodes (Occurrence Threshold)
over a maximum length of time (Event Window).

The Coincident Event Removal drop-down menu specifies how AxIS will search for coincident spikes across
the plate. Select from:

1. None: No search will be performed.

[14]
 AxIS Overview

2. Well: Search will look for coincidence across electrodes in a well.

3. Plate: Search will look for coincidence across electrodes in any well on the plate.
4. Chip: Search will look for coincidence across electrodes that share a hardware circuitry connection.
For example, if Coincident Event Removal is set to Well with Occurrence Threshold equal to 4 electrodes and
Event Window equal to 80 µs, spikes that occur on at least 4 electrodes in the same well within 80 µs of each
other will be ignored.

The Spike Counting box sets the bin time for counting spikes as defined by the Interval field. The bin time is
used by the Activity Map and Statistics Graphing modules and the Spike Count files.

2.3.6. Burst Detector

The Burst Detector is a neural data processor that analyzes spike timing patterns to identify bursts on individual
electrodes (single-electrode bursts) and across multiple electrodes (network bursts). See Section 7.3 for more
information about neural bursting. Because it requires spike data, a Burst Detector can only be placed after a
Spike Detector. Neural Statistics Compiler data processors (Section 2.3.9) may only be placed below a Burst
Detector. The Burst Detector has two outputs: Electrode Burst List (.csv) and Network Burst List (.csv).

Burst Detector results are tracked by the raster plot in the Spike Plots module. Spikes belonging to single-
electrode bursts are plotted on the raster in the same color as the Burst Detector plot color. Network bursts are
indicated by a box of the contrasting color.

The Burst Detector has two single-electrode burst detection methods selectable in the Single Electrode Bursts
drop-down menu:

1. Inter-Spike Interval (ISI) Threshold: An electrode burst is defined as at least N spikes on an electrode,
each separated by an inter-spike interval (ISI) of no more than T seconds. The method is adapted

[15]
AxIS 2.4 User Guide

from Chiappalone et al., 2005. The minimum number of spikes and maximum time between each
spike are set using the Minimum Number of Spikes and Maximum Inter-Spike Interval fields,
respectively.
2. Poisson Surprise: This algorithm assumes the neurons are firing according to a Poisson distribution. A
collection of spikes is compared to the probability that this collection of spikes would occur by
chance. If the collection of spikes exceeds the “surprise” threshold, then it is considered an electrode
burst. The Minimum Surprise field determines how sensitive burst detection is. A higher threshold
makes the burst detector less likely to detect an electrode burst, while a lower threshold is more lenient
and more likely to accept a less “surprising” collection of spikes as burst. The method is adapted from
Legéndy & Salcman, 1985. In this way, the algorithm is adaptive to the mean firing rate on each
electrode, computed dynamically to enable real-time detection and visualization of bursts.

Use the checkbox in the Network Bursts section to Enable Network Burst Detection. As in the single-electrode
burst definition, the network burst algorithm relies on defining a network burst as a collection of N spikes, each
separated by an inter-spike interval of no more than T seconds, however the spikes are detected across the
well instead of a single electrode. In addition to the above criteria, a minimum number of electrodes (E) must
be contributing to the network burst. In the Network Bursts box, set N spikes with Minimum Number of Spikes,
T seconds with Maximum Inter-Spike Interval, and E electrodes as a percentage of the total electrodes with
Minimum Participating Electrodes (%).

The Detection Window field in the Mean Firing Rate Estimation box sets the duration of the sliding window
used to calculate the mean firing rate for Poisson Surprise burst detection.

The Window Size field in the Synchrony Parameters box is the window of time around zero phase-lag used to
compute the area under the cross-correlation and area under the normalized cross-correlation synchrony
metrics provided by AxIS. See Section 7.3 for further explanation of synchrony.

2.3.7. Cardiac Beat Detector

The Cardiac Beat Detector processor detects threshold crossings in the continuous data stream to identify
“beats” and calculates associated cardiac endpoints. The beats are plotted in the Cardiac Beats Plots module
as cardiac waveform, conduction, and beat period plots. The Cardiac Beat Detector is the base processor for
cardiac data visualization and analysis. It identifies the beat timing, propagation, amplitude, and duration and
is used as a source for the Cardiac Beat Plots, Activity Map, and Statistics Graphing modules. Cardiac
Statistics Compiler data processors (Section 2.3.8) may only be placed below a Cardiac Beat Detector. The
Cardiac Beat Detector has two outputs: Electrode Beat List (.csv) and Well Beat List (.csv). See Chapter 6 for
more information about cardiac analysis endpoints.

[16]
 AxIS Overview

The Cardiac Beat Detector settings are divided into three sections: Beat Detection Parameters, Conduction
Parameters, and Display.

Setting a proper detection threshold is crucial for accurate cardiac data analysis. Beat detection is controlled
by three fields:

1. Detection Threshold: Sets the threshold for detecting a beat. When the continuous voltage data
exceeds this threshold, the algorithm begins searching for a beat that meets the Min Beat Period and
Max Beat Period parameters. If a beat is identified, the beat time is marked as the point of maximum
slope of the depolarization spike. The threshold should be lower than the initial depolarization spike
amplitude and higher than any other feature of the cardiac waveform (300-600 µV is
recommended).
2. Min Beat Period: The minimum time between two threshold crossings to be considered a beat.
3. Max Beat Period: The maximum time between two threshold crossings to be considered a beat.

Field potential duration (FPD) is calculated as the time between the depolarization and repolarization, noted
by the beat time and the repolarization peak or t-wave, respectively. Due to processor constraints, FPD
detection is disabled during data acquisition. To identify the t-wave, there are three methods selectable in the
FPD Method drop-down menu:

[17]
AxIS 2.4 User Guide

1. Polynomial Regression (Recommended): Performs a polynomial regression to identify a peak or

trough between two time points. The search window starts at the Post Spike Detection Holdoff, the
time after the current depolarization spike. The window ends with either the Pre Spike Detection
Holdoff, the time before the next depolarization spike, or the Max Post Search Duration, a fixed
maximum time after the current depolarization spike. The T-wave Detection Feature drop-down
determines whether the regression searches for a peak (Max), trough (Min), or either (Auto
(Max/Min)).

2. Inflection Search: Segments the search window to identify the region containing the t-wave and then
performs a polynomial regression on that segment to determine the exact location. To identify the t-
wave segment, the algorithm searches for a region that crosses a threshold X times the noise as set by
Detection Threshold. After identifying the segment, a polynomial regression is performed on a
window of size Regression Window Size. The search window starts with the Post Spike Detection
Holdoff, the time after the current depolarization spike, and ends with the Pre Spike Detection
Holdoff, the time before the next depolarization spike. The number of segments is set by the Beat
Segmentation field. The T-wave Detection Feature drop-down determines whether the regression
searches for a peak (Max), trough (Min), either (Auto (Max/Min)), or the maximum slope of the
waveform (dV/dt).

3. Zero Crossing: Performs a polynomial regression to identify a zero crossing between two time points.
The search window starts with the Post Spike Detection Holdoff, the time after the current
depolarization spike. The window ends with either the Pre Spike Detection Holdoff, the time before
the next depolarization spike, or the Max Post Search Duration, a fixed maximum time after the
current depolarization spike. The first zero crossing of this regression is chosen as the t-wave location.
Averaging multiple cardiac waveforms across beats can reduce noise and increase t-wave resolution and
recognition by AxIS. The cardiac waveform displayed in the Cardiac Beat Plots module is the average of the
current beat and the previous N-1 beats as set by the Running Average Beat Count. Use the averaged beat
for FPD detection by enabling the Use for FPD Detection checkbox.

The Conduction Parameters section identifies well beats, also called synchronized beats. A synchronized beat
is a depolarization spike detected on a minimum number of electrodes within a well in a certain time window.
The synchronized beat time is marked as the beat time of the first participating electrode. The minimum
percentage of participating electrodes is set by Min Active Electrodes. The time frame is set by Max
Propagation Delay. Synchronized beats are displayed on the Conduction plot in the Cardiac Beat Plots
module.

The Display section controls the data visualization in the Cardiac Beat Plots module. It has the following
settings:

1. Display Start: The amount of time displayed before the depolarization spike.
2. Display End: The amount of time displayed after the depolarization spike.
3. Averaged Beats checkbox: If enabled, displays the averaged beat waveform plot instead of the
instantaneous beat waveform plot for each electrode. The number of beats used to create the
average is specified in the Running Average Beat Count field in Beat Detection Parameters.

[18]
 AxIS Overview

4. Synchronized Beats Only checkbox: If enabled, displays only beats that qualify as synchronized
beats.
5. Scale Beat Brightness checkbox: If enabled, beats with large amplitude depolarization spikes are
displayed brighter than beats with low amplitude depolarization spikes.
6. Lowest Intensity: If Scale Beat Brightness is enabled, sets the amplitude of low amplitude
depolarization spikes.
7. Highest Intensity: If Scale Beat Brightness is enabled, sets the amplitude of large amplitude
depolarization spikes.
8. FPD Marker checkbox: If enabled, displays a rectangle marking the current T-wave location,
providing a visual verification of the FPD detection accuracy.
9. FPD Confidence Interval checkbox: If enabled, displays a white whisker plot on the cardiac waveform
plot indicating the mean T-wave location and the confidence interval as set by Confidence Interval
Markers (x Std Dev).
10. Confidence Interval Markers (x Std Dev): Sets the width of the confidence interval displayed as a
multiple of the standard deviation.

The checkbox controls can be set by right-clicking on the beat waveform plots in the Cardiac Beat Plots
module, and selecting/deselecting the display options.

2.3.8. Cardiac Statistics Compiler

The Cardiac Statistics Compiler (displayed as Statistics Compiler in the Streams window) uses the outputs of a
Cardiac Beat Detector to calculate a variety of cardiac endpoints. These metrics are calculated for individual
electrodes, well-wide averages, and treatment group averages. Treatment groups are defined by the Plate
Map Editor in the Active Plate interface. The Cardiac Statistics Compiler can only be used after a Cardiac
Beat Detector and only on recorded data. It has one output file, Advanced Metrics (.csv). See Section 5.2 for
more information on file types.

[19]
AxIS 2.4 User Guide

The settings window has three sections, Saved Data, Beat Quality Control, and FPD Measure Quality Control.
The Saved Data section identifies what information is output to the Advanced Metrics file. The options include:

1. File Header: If enabled, output file will start with a list of all experiment acquisition and analysis
settings.
2. Aggregated Well Statistics: If enabled, output file will include well-wide averages in addition to
individual electrode metrics.
3. Aggregated Treatment Statistics: If enabled, output file will include treatment group averages as
defined by the Plate Map Editor.
4. Source Data: If enabled, output file will include lists of individual beat statistics for both synchronized
beats and electrode beats.
5. Supplemental Statistics: If enabled, additional cardiac endpoints are included in the file. Examples
include median and median absolute deviation (MAD) for each metric and additional conduction
velocity metrics.

The Beat Quality Control section limits which beats are averaged and reported. The options include:

1. Limit to Region of Most Stable Beat Period checkbox: If enabled, metrics will be calculated from the
most stable well beats, calculated as N consecutive well beats with the lowest beat period standard
deviation. If fewer than N beats are detected in the file, all detected beats will be used to calculate
metrics.
2. Number of Beats In Stable Region: Defines the number of consecutive well beats used for the Limit to
Region of Most Stable Beat Period checkbox.
3. Included Source Data: Specifies whether to output the source data for all well beats in the recording
(All Beats) or for the most stable well beats only (Most Stable Region Only).
Note: All Beats is required for the CiPA Analysis Tool.
4. Region Limit: Sets the time window of the recording used to search for the most stable beats. Specify
the time window by selecting Entire Playback, Start of Playback, or End of Playback, to choose the
entire recording, the beginning with or without an offset, or the end of the recording, respectively.
5. Limit Statistics to Most Common Propagation Pattern checkbox: If enabled, metrics will be calculated
from well beats that follow the most common propagation pattern.

The FPD Measure Quality Control section removes FPD measurements from beats and electrodes that do not
meet user-specified statistical bounds, improving the accuracy of electrode and well-wide FPD averages. The
number of remaining electrodes that are used for the well FPD calculation after the criteria are applied is
output as “Total FPD Electrodes”. The user-specified statistical bounds are applied in the following order:

1. Beat to Beat FPD Consistency: Individual beat metrics are compared to the mean or median of all
beats on that electrode. FPD values for individual beats are removed when they exceed the specified
number of standard deviations (x STD) or median absolute deviations (x MAD) from the mean or
median, respectively.

[20]
 AxIS Overview

2. Electrode FPD Consistency: Electrodes with large beat-to-beat variability in FPD are removed.
Consistency is based on the coefficient of variation (CoV) of the FPD, with limits set as a percentage
([standard deviation/mean] *100).
3. Well FPD Consistency: Electrodes with an FPD mean that exceeds the specified number of standard
deviations (x STD) or median absolute deviations (x MAD) from the well mean or median are
excluded from the Well Average FPD.

2.3.9. Neural Statistics Compiler

The Neural Statistics Compiler (displayed as Statistics Compiler in the Streams window) uses the outputs of a
Spike Detector and Burst Detector processor to calculate a variety of spike, burst, and synchrony metrics.
These metrics are calculated for individual electrodes, well-wide averages, and treatment group averages.
Treatment groups are defined by the Plate Map Editor in the Active Plate interface. The Neural Statistics
Compiler can only be used after a Burst Detector and only on previously recorded data. It has one output file,
Advanced Metrics (.csv). See Section 5.2 for more information on file types.

The settings window identifies what information is output to the Advanced Metrics file. The options include:

1. File Header: If enabled, output file will start with a list of all experiment acquisition and analysis
settings.
2. Aggregated Well Statistics: If enabled, output file will include well-wide averages in addition to
individual electrode values.
3. Aggregated Treatment Statistics: If enabled, output file will include treatment group averages as
defined by the Plate Map Editor.
4. Synchrony: If enabled, output file will include synchrony metrics.

Active Electrode Selection Criteria: Identifies active electrodes defined as having a mean firing rate greater
than the value defined in the Minimum spike rate field. The output file reports the number of active electrodes
in each well, as well as the weighted mean firing rate, the mean firing rate averaged across only the active
electrodes.

[21]
AxIS 2.4 User Guide

2.4. FILE PLAY AND DISPLAY CONTROLS

The File Play and Display controls activate the current stream and control data display in the active window.

The File Play options are:

Control Icon Description

Load/Eject Plate Engages or disengages a plate from the Maestro dock. For
Maestro systems with automated docking only.
Stop Stops the current stream. Followed by a Play or Record
command will resume streaming a .raw file from the start. A live
stream will resume with a voltage offset.
Pause Pauses the current stream. Followed by a Play or Record
command will resume a .raw file at the time it was paused. A
live stream will resume without a voltage offset.
Play Starts the current stream without saving. Data will be displayed
in the Control Bar modules.
Record Records and plays the current stream. Data will be displayed in
the Control Bar modules and outputs selected in Experiment
Setup Properties will be saved.

The Display controls scale data displayed in the active window and, in certain modules, provides a drop-
down menu to select the type of data displayed. Use the Source drop-down to select the type of data
displayed. Use the zoom in ( ), zoom out ( ), and reset scale ( ) controls to change the voltage and time
scales. The scale is indicated above the control as X units per division. The divisions are noted along the edges
of the active window.

2.5. STATUS BAR

The Status Bar at the bottom of the active window displays the file status of the current data stream. The time a
stream has played for is indicated on the left. The duration of a recording is indicated on the right. The text in
the middle of the bar includes the Stream name, play status, well, and treatment information if present in the
Active Plate. When the active stream is a previously recorded file, the status bar represents the time within the
file. Click on the status bar to move to a particular time in the recording. The color of the status bar indicates
the current file status as outlined in the table below.

Stream Play Time Recording Duration

[22]
 AxIS Overview

Color File Status

Light Gray Current stream is inactive, or live stream is active and playing
Blue Gray Offset correction in progress.
Green Current stream is a recording. Dark green indicates time before the
currently displayed data.
Red Current stream is recording. Output files specified in Experiment Setup
Properties are being saved.

2.5.1. Adding Timestamp Notes

It is possible to add notes with a timestamp to a recording. Notes can mark important events for further review
during analysis. Notes may be added while the file is playing, paused, or recording. A line will appear in the
Status Bar at the time of the notes addition. When selected or scrolled over, the note will display its message.

To add a timestamp note:

1. Click Add Note located to the right of the Status Bar at the desired file time.
2. Type the desired text and click Accept (green checkmark).
3. Click the in the top right corner to close the Notes panel.
Note: Notes are only available in files recorded with ‘AxIS’ versions 1.7.9 or higher. To enable notes in data
recorded with previous versions, record a copy of the .raw file again with a newer version of ‘AxIS’. See
Section 5.4.1.

Accept,
Cancel,
Delete

Note Location Add Note

[23]
AxIS 2.4 User Guide

2.6. CONTROL BAR

The Control Bar interface switches the module displayed in the active window. By default, only modules used
by the current configuration in the Streams window are displayed. To display all Control Bar modules, select
Show all Modules under the View menu. To switch the active module, click on the thumbnail.

Experiment Stimulation Statistics Continuous Cardiac

Setup Studio Graphing Plots Beat Plots

Scheduled Environment Activity Spike

Recording Control Map Plots

Name Description Dependency

1. Experiment Setup Properties File information controls. Sets save location,
experiment notes, recording names, and file Always visible
types.
2. Scheduled Recording Setup Automated recording and stimulation
Live data only
controls.
3. Stimulation Studio Stimulation definition controls. Always visible
4. Environmental Control Temperature and CO2 controls. Live data only
5. Statistics Graphing Live display of stream-based statistics. Data stream active
6. Activity Map Plate-wide visualization of spike or beat rate Spike Detector
and amplitude. Cardiac Beat Detector
7. Continuous Waveform Plots Displays raw and/or filtered voltage over
time plot for every electrode in a selected Data stream active
well.
8. Spike Plots Displays spike waveforms for every
electrode in a selected well and well raster Spike Detector
plot.
9. Cardiac Beat Plots Displays cardiac waveforms, conduction
maps, and beat period plots for every Cardiac Beat Detector
electrode in a selected well.

2.6.1. Experiment Setup Properties

The Experiment Setup Properties module specifies the recording destination, experiment description, and the
file names and types to record. Available recording file types (see Section 5.2) are determined by the data
processors present in the Streams window.

[24]
 AxIS Overview

Set Recording Destination

Experiment Details

File Name
File Type

Experiment Setup

The Investigator, Experiment ID, and Description fields are optional fields for the user to enter information
about the experiment. This information will be saved with the file and appear on the analysis output files. They
may be entered or updated at any time. Investigator and Experiment ID can be applied to file names as
macros (Section 5.2.1).

To update the Investigator, Experiment ID, or Description fields on a previously recorded file:

1. Click on the Experiment Setup Properties module in the Control Bar.

2. Click on the field and enter the new information
3. Click Write to File.
Note: To undo changes prior to clicking Write to File, click Reset.

To set the file save location:

1. Click on the Experiment Setup Properties module in the Control Bar.

2. Click the Browse… button.
3. Navigate to and select the destination folder.
Note: When analyzing previously recorded data, analysis files generated are saved to the folder of
the original data. This destination cannot be changed in ‘AxIS’.

2.6.2. Scheduled Recording Setup

The Scheduled Recording Setup module automates data recording and stimulation. The module is divided into
two sections: Settings and Status.

[25]
AxIS 2.4 User Guide

Settings

Status
Start Schedule

Scheduled Recording

The Settings dialog controls how and when a file is recorded.

Field Description Options

Record Every The interval between the start of Seconds, Minutes, Hours, or Days
each recording
Record For The duration of each recording Seconds, Minutes, Hours, or Days
Starting Immediately – begin when Start Schedule is
clicked.
At – begin at a specified time (mm/dd/yyyy
The start time of the first
hh:mm:ss).
recording.
Note: 24 hour time or AM/PM may be used. If
starting Immediately, play the stream first to allow
voltage offset to complete, then Start Schedule.
Execute Until Stopped – continue at set interval until Stop
Schedule is clicked.
Once – Only create one recording.
Number of repeats
Until – continue at set interval until a specified
time.
Exactly – repeat at set interval for X times.
Auto Stimulate Starts the currently active From Recording Start – When to start the
stimulation protocol during a stimulation protocol after the start of a recording if
scheduled recording. Auto Stimulate is selected.
Note: Only available if
stimulation protocol is set to
Once in Stimulation Studio.

After specifying the Settings, click Start Schedule to begin the scheduled recording protocol.

[26]
 AxIS Overview

The example below shows the recording and stimulation times for a Scheduled Recording Setup as follows:

Record Every: 15 Minutes

Record For: 6 Minutes
Starting: At 8:00 AM
Execute: Exactly 4 times
Auto Stimulate: Enabled, 3 Min From Recording Start

8:00:00 AM 8:15:00 AM 8:30:00 AM 8:45:00 AM 9:00:00 AM

Record
Stimulate

The Status panel provides feedback about the recording schedule.

Field Description
Running Status Current recording status. Either:
Recording – a scheduled recording is in progress.
Waiting to Record – another recording is scheduled to begin
Waiting to Trigger – a scheduled recording is in progress and a stimulation is
scheduled to begin.
Next Recording Start time of the next scheduled recording
Previous Recording Start time of the last completed recording

2.6.3. Stimulation Studio

The Stimulation Studio module designs both electrical and optical stimulation protocols. The module has four
key components: the stimulation blocks, stimulation lanes, global repeat settings, compiled pattern view, and
electrode or LED selector. See Chapter 4 for more information about building stimulation waveforms, selecting
stimulation electrodes or LEDs, and applying a stimulus.

Stimulation Blocks Compiled Pattern View

Global Repeat Settings

Stimulation Lanes

Electrode or
LED Selector

Stimulation Studio

[27]
AxIS 2.4 User Guide

2.6.4. Environmental Control

The Environment Control module displays current temperature and, when integrated with an Axion Gas Mixer,
CO2 levels. The Environmental Control thumbnail displays the current temperature and CO2 status. The color
of the text indicates the current status.

Name Status Description

Gray Text Maestro is connected and the heater or gas mixer is off.

Maestro is connected, heater or gas mixer is on, and

Blue Text
temperature or CO2 has not yet reached the set point.

Maestro is connected, heater or gas mixer is on, and

Green Text
temperature or CO2 is at the set point.

Yellow Text Temperature or CO2 is slightly above the set point.

Red Text Temperature or CO2 is greatly above the set point.

Navigate to the Environment Control module using the Control Bar. The active window will display as below:

Heater Settings

CO2 Settings

[28]
 AxIS Overview

The temperature and CO2 plots display the current set point as a dashed line and the current temperature or
CO2 as a solid line. To the right of the temperature plot are the heater controls, including the Heater Control
switch and Temperature (°C) set point. The CO2 control box contains the CO2 Control switch, set point, gas
flow rate, and CO2 alarm settings. If enabled, the CO2 alarm will display a warning when the CO2
concentration is outside of the Max Deviation (%) away from the set point (%CO2) for longer than the Max
Deviation Time (min).

To set the temperature:

1. Click on the Environmental Control module in the Control Bar.

2. Set the desired temperature in the Temperature (°C) field (range 0-46°C).
Note: Temperatures above 37°C may be detrimental to cells.
3. Click the Heater Control switch to On.
The heater can take several minutes to arrive at its set point; turn it on before a live experiment begins.

To set the CO2 concentration:

1. Click on the Environmental Control module in the Control Bar.

2. Set the desired CO2 concentration in the Concentration (%) field and flow rate in the Flow Rate
(l/min) field.
3. Optional: Set the Max Deviation (%) and Max Deviation Time (min) and turn the Alarm Settings
switch to On.
4. Click the CO2 Control switch to On.

2.6.5. Statistics Graphing

The Statistics Graphing module displays metrics plotted over time on a per-well basis for the entire plate. The
dotted line represents the well median, the solid line the well mean. The metric displayed is selected by the
Source drop-down menu. The options available depend on the data processors in the Streams window as
listed below:

Data Processor Source Metric

Raw Data Standard deviation of the voltage.
Filter Standard deviation of the filtered voltage.
Cardiac Beat Detector Spike Amplitude
Beat Rate
Field Potential Duration
Spike Detector Spike Rate
Spike Amplitude
Burst Detector Burst Rate
Burst Duration
Synchrony Index
Network Burst Duration
Network Burst Rate

[29]
AxIS 2.4 User Guide

Source Data No Detected Spikes

Statistics Graphing

2.6.6. Activity Map

The Activity Map module displays a heat map of activity across the entire plate, providing an intuitive way to
visualize differences between wells/conditions.

Source

Scale

Activity Map

The metric displayed is selected by the Source drop-down menu. The options available depend on the data
processors in the Streams window as listed below:

[30]
 AxIS Overview

Data Processor Source Metric

Cardiac Beat Detector Spike Amplitude (µV)
Beat Rate (beats/minute)
Spike Detector Spike Amplitude (µV)
Spike Rate (Spikes/second)

The refresh rate of the Activity Map is set by the Interval setting in the Spike Detector (See Section 2.3.5).
Small intervals increase the refresh rate, providing real-time feedback. Larger intervals will display a persistent
signal integrated over many spikes.

Color corresponds to the magnitude; red or white areas have the highest magnitudes while blue and black
areas have the lowest. The Activity Map scale is located on the bottom right of the active window. The scale
auto-adjusts by default but may be set to a fixed value in the right-click menu.

To manually adjust the Activity Map scale:

1. Click on the Activity Map module in the Control Bar.

2. Right-click on the scale color bar.
3. Deselect Auto Scale if enabled.
4. Enter the desired scale maximum into the Maximum field.
5. Click Enter.

To copy the Activity Map:

1. Click on the Activity Map module in the Control Bar.

2. Right-click on the Activity Map and select Copy.

Selecting a well in the Activity Map will change the active well displayed in the Continuous Waveform Plots,
Spike Plots, and Cardiac Beat Plots modules.

2.6.7. Continuous Waveform Plots

The Continuous Waveform Plots module displays continuous voltage data for each electrode of the active
well. Select the active well by clicking on the desired well on the Active Plate or Activity Map.

[31]
AxIS 2.4 User Guide

Electrode ID Noise

Active Well

Continuous Waveform Plots

The electrode ID consists of two digits, the column and row. Adjacent to the electrode ID is the electrode noise
level, displayed in µV.

The x and y scale is adjusted in the Display controls. Division marks along the outside of the active window
indicate the zoom level chosen in the Display controls section. See Section 2.4 for more information about the
Display controls. By default, electrodes are laid out in the active window by physical location but may be
viewed as a vertical stack.

To view the electrodes in a vertical stack:

1. Click on the Continuous Waveform Plots module in the Control Bar.

2. Right-click on the active window and select Stack Plots.

The continuous waveform plot has two copy options. To copy a single electrode, right-click on the electrode
and select Copy Electrode Waveform. To copy the entire well plot, right-click anywhere on the plot and select
Copy Well Plot.

A blank plot indicates an electrode has been turned off. Right click on an electrode plot and select Disable
Electrode to turn it off or on, or use the Plate Map Editor (Section 2.2.2).

2.6.8. Spike Plots

The Spike Plots module uses output from a Spike Detector and Burst Detector (See Sections 2.3.5 and 2.3.6,
respectively) to display the spike voltage waveforms and raster plot of the active well. Like the Continuous
Waveform Plots, each electrode of the active well is represented in the active window of the Spike Plots

[32]
 AxIS Overview

display. The gray electrode ID corresponds to the electrode ID in the Continuous Waveform Plots and the
adjacent number is the most recent spike amplitude.

Electrode ID Spike Amplitude

Display Controls

Raster Plot Raster Zoom

Spike Plots

The waveform plots overlay each spike with the brightest trace representing the most recently detected spike.
Previously detected spikes fade to black. Empty black panels indicate no spikes have been detected on those
electrodes. Note the dedicated stimulation electrode on E-Stim+ plates is left empty. Spike detection and
display settings can be adjusted in the Spike Detector Settings (See Section 2.3.5).

The x and y scale of the waveform plots are adjusted in the Display controls. Division marks along the outside
of the active window indicate the zoom level chosen in the Display controls section. See Section 2.4 for more
information about the Display controls. By default, electrodes are laid out in the active window by physical
location.

The waveform plot has two copy options. To copy the entire well image, right-click anywhere on the plot and
select Copy Well Plot. To copy a single electrode, right-click on the electrode and select Copy Electrode
Waveform. This copies both the electrode image as well as the time and voltage values used to generate the
image. To paste the electrode image, right-click and select Paste Special and choose the image option. When
pasting the data samples, the first column contains the time in seconds, and the remaining columns list the
voltage values associated with each time sample for the displayed spikes. Each column contains the data
samples for a single spike waveform, up to the maximum of 20 spikes displayed at any given time, with the
most recent spike data in the last column.

[33]
AxIS 2.4 User Guide

The raster plot shows a running display of marks indicating the time each spike occurred with each row on the
plot displaying the spikes from a single electrode. Single-electrode bursts are displayed in the color indicated
by the Burst Detector while network bursts are indicated by a box of the contrasting color. Right-click to
enable/disable display of network bursts. The time scale is controlled by the zoom controls on the right,
allowing monitoring of spike history over long timescales.

The raster plot can be copied as an image. To copy the raster plot, right-click on the raster plot and select
Copy.

2.6.9. Cardiac Beat Plots

The Cardiac Beat Plots module uses output from a Cardiac Beat Detector (See Section 2.3.7) to display the
cardiac voltage waveforms, conduction maps, and beat period plots of the active well. Like the Continuous
Waveform Plots, each electrode of the active well is represented in the active window of the Cardiac Beat
Plots display. The gray electrode ID corresponds to the electrode ID in the Continuous Waveform Plots and the
adjacent number is the average amplitude of the depolarization spike, displayed in µV.

Spike Amplitude

T-wave Propagation Map

Beat Period Plots

Right-click Menu

Cardiac Beat Plots

By default the average waveform of the last 10 beats are plotted. This may be changed either in the Beat
Detector Settings (section 2.3.7) or by right-clicking on the plot and selecting/deselecting Averaged Beats.
The field potential duration (FPD) detection is indicated by the blue rectangle and white whisker plot. It is
disabled during data acquisition to limit processor workload.

The x and y scale is adjusted in the Display controls. Division marks along the outside of the active window
indicate the zoom level chosen in the Display controls section. See Section 2.4 for more information about the

[34]
 AxIS Overview

Display controls. By default, electrodes are laid out in the active window by physical location but may be
viewed as a vertical stack.

To view the electrodes in a vertical stack:

1. Click on the Cardiac Beat Plots module in the Control Bar.

2. Right-click on the active window and select Stack Plots.

The waveform plots have two copy options. To copy a single electrode, right-click on the electrode and select
Copy Electrode Waveform. This copies both the electrode image as well as the time and voltage values used
to generate the image. When pasting the data samples, the first column contains the time samples in seconds
and the second column contains the voltage samples corresponding to the selected beat. To paste the
electrode image, right-click to select Paste Special and choose the image option. To copy the entire well
image, right-click anywhere on the plot and select Copy Well Plot.

To the right of the cardiac waveforms are three plots: the Ending Electrode
Conduction plot, Beat Periods plot, and Beat Period Poincaré plot.

Right-click on any of the plots and select Copy to copy to the

clipboard.

The Conduction Plot illustrates the propagation delay at each

electrode from short (blue) to long (red). The well beat originates in
the blue region and terminates in the red. The electrode column and No Beat on this Electrode
row numbers are displayed on the x- and y-axis, respectively.

Starting Electrode
Right-clicking on the Conduction plot presents four display options:

Option Description
Auto Scale Sets the scale of the plot. Disable to manually adjust the top of the
scale with Max Delay
Animate Conduction Animates the plot with each beat. Disable to view a static plot that
updates with each beat.
Extrapolate Missing Electrodes Extrapolates values for electrodes that did not detect a
depolarization spike.
Show Missing Electrodes Extrapolated electrodes will be marked with a black circle.

[35]
AxIS 2.4 User Guide

The Beat Periods plot displays beat period (in

seconds) on the y-axis, plotted against beat
number on the x-axis. The Beat Period
Poincaré plot displays beat period (in
seconds) on the y-axis, plotted against the
previous beat period on the x-axis (BPt+1 vs.
BPt). Beats that exhibit a greater than 5%
deviation from the previous beat are
displayed as red, indicative of arrhythmia.

Below the Beat Period Poincaré plot are

display controls for both plots. Use the zoom
in ( ), zoom out ( ), and reset scale ( )
controls to change time scales. The scale is
indicated above the control as X units per
division and N beats.

[36]
 Data Acquisition

CHAPTER 3. DATA ACQUISITION

The Maestro and Muse stream data from all electrodes simultaneously. AxIS can record the continuous
voltage data from the entire plate at once, and the user may view the data stream with various data
processors applied. Built-in environmental controls ensure the culture is kept under optimal conditions while
recording. This chapter reviews how to acquire data from the Maestro/Muse using AxIS. Section 3.3 contains
a step-by-step tutorial for data acquisition.

3.1. AXIS COMMUNICATION WITH THE MAESTRO/MUSE

AxIS communicates with a Muse through a USB connection and with the Maestro through USB and Ethernet
connections. Once the system is properly connected, turn on the Middleman using the switch on the back and
open AxIS to begin system initialization. Startup can take a few minutes.

The status lights in the bottom right corner of AxIS indicate the system status and connectivity. There are three
status lights: MEA Status, Lumos Status, and Remote Control Status.

Maestro/Muse Connection Status

Device not detected

Blinking. Connected to Device, and Device is booting

up

Connected to Device, no MEA plate detected

Connected to Device, MEA plate detected

Error

Lumos Connection Status

Lumos not detected

Blinking. Connected to Lumos, and Lumos is booting

up
Connected to Lumos, and Lumos is on the stand

Connected to Lumos, and Lumos is on the Maestro

Error

Remote Control Status

Remote Control is not enabled

Remote Control is enabled

[37]
AxIS 2.4 User Guide

3.2. CONFIGURING THE MAESTRO/MUSE FOR ACQUISITION

3.2.1. Configuring the Hardware

The Maestro or Muse must be configured prior to recording a continuous voltage stream. The hardware
bandwidth and gain must be properly set for particular applications. These settings affect how the raw
continuous voltage data is recorded, and cannot be changed after data is recorded. Software filtering added
by a Digital Filter data processor may be changed after recording (Section 2.3.2).

Axion provides preset Neural or Cardiac recording configurations that will automatically configure the
recommended hardware settings (Section 3.2.2).

With a device connected, use the Settings dialog from the Maestro or Muse stream to adjust the analog
acquisition settings. To access the Settings dialog, double-click on the Maestro or Muse stream, or right-click
and select Settings.

The Maestro or Muse Settings contain the following options:

1. Sampling Frequency: The rate data is sampled and collected. For the Maestro, the rate is fixed at
12.5 kHz. Muse users may select from 12.5 kHz or 25 kHz. A higher sampling frequency will
increase the file size.
2. Analog Settings: Configure the hardware bandwidth and gain. There are 7 options for various neural
and cardiac applications:
Option Setting Description
Neural: Spikes Gain: 1200X High gain with a neural bandwidth.
(default neural) Bandwidth: 200-5000 Hz Recommended for most neural
applications.
Neural: Broadband Gain: 1200X High gain with wide neural bandwidth.
Bandwidth: 10-5000 Hz Use if a wider bandwidth is desired.
Field Potentials Gain: 1200X High gain with lower frequency
Bandwidth: 1-2000 Hz bandwidth.

[38]
 Data Acquisition

Cardiac: Low Amplitude Gain: 1200X High gain with wide cardiac bandwidth.
Bandwidth: 0.1-2000 Hz Suitable for signals too small to detect
using normal cardiac settings.
Cardiac: Pacing Gain: 1200X High gain with a cardiac bandwidth.
Bandwidth: 1-2000 Hz Low pass filter cutoff frequency is lower
to reduce stimulation artifact. Use for
cardiac recordings with electrical
stimulation.
Cardiac: Standard Gain: 130X Low gain with a cardiac bandwidth.
(default cardiac) Bandwidth: 1-25000 Hz Recommended for most cardiac
applications.
Cardiac: Broadband Gain: 130X Low gain with a wide cardiac
Bandwidth: 0.1-25000 Hz bandwidth. Use if a wider bandwidth is
desired.

3. Low Pass Filter drop-down: Applies a 2 kHz or 3 kHz Kaiser Window filter. A 2 kHz Kaiser Window
filter is recommended for cardiac recording. When a cardiac option is selected in Analog Settings,
the 2 kHz filter is enabled by default. It is disabled when a neural option is selected.
4. Referencing drop-down: Subtracts the mean or median signal from groups of electrodes. Grouping is
performed based on the system’s electronic architecture. This reduces the noise common to the
electrode groups, improving the detection of low amplitude signals. Median referencing is
recommended for neural recording. When a neural option is selected in Analog Settings, median
referencing is enabled. It is disabled when a cardiac option is selected.

3.2.2. Stream Configurations

In addition to hardware configuration, it is possible to add data processors for easy data visualization during
a recording. These additional data processors do not impact the raw continuous voltage data recorded and
are purely for visualization.

A stream configuration sets the hardware configuration, adds data processors, and applies all the settings for
data visualization and acquisition.

AxIS comes with a variety of preset configurations. Use the Real-Time configurations for data acquisition:

Configuration Processing Applied Description

Cardiac Real-Time
Spontaneous Digital Filter Generates activity map, cardiac waveforms, beat rate
Cardiac Beat Detector and conduction information. FPD detection disabled.

Electrically Artifact Eliminator Generates activity map, cardiac waveforms, beat rate
Paced Cardiac Beat Detector and conduction information. Optimized for reducing
stimulus artifacts in pacing experiments. FPD detection
disabled.

[39]
AxIS 2.4 User Guide

Optically Paced Digital Filter Generates activity map, cardiac waveforms, beat rate
Cardiac Beat Detector and conduction information. FPD detection disabled.

Neural Real-Time
Spontaneous Digital Filter Generates activity map, spike waveforms, and raster
Spike Detector plot. Network burst detection disabled.
Burst Detector
Electrically Digital Filter Generates activity map, spike waveforms, and raster
Evoked Artifact Eliminator plot. Optimized for reducing stimulus artifacts in
Spike Detector electrically-evoked experiments. Network burst
Burst Detector detection disabled.
Optically Evoked Filter Generates activity map, spike waveforms, and raster
Spike Detector plot. Network burst detection disabled.
Burst Detector

To apply a configuration:

1. Right-click on the data stream.

2. Select Configuration and navigate to the desired configuration. Click on the configuration.
Note: A configuration adds data processors but also sets hardware configurations. To record with
non-default analog settings, apply the configuration first and then manually change the analog
settings. See Section 3.2.1.

To save a custom configuration:

1. Manually set up a data stream with desired hardware configurations, data processors, and selected
file outputs.
2. Right-click on the data stream.
3. Select Configuration  Save.
4. Type a filename (.datastreams extension) in the Save Data Stream Configuration dialog and click
Save.

To load a custom configuration:

1. Right-click on the data stream.

2. Select Configuration  Load.
3. Navigate to the .datastreams configuration file in the Open Data Stream Configuration dialog.
4. Click Open.

[40]
 Data Acquisition

3.3. LIVE DATA ACQUISITION TUTORIAL

While most data processor and plate layout settings can be modified after a recording has taken place there
are four key settings that must be correct prior to recording:

1. The Maestro or Muse must be communicating with AxIS. See Section 3.1.
2. The environment control must be on and equilibrated. See Section 2.6.4.
3. The hardware configuration must be set. See Section 3.2.
4. Any stimulus applied must be configured before starting the stimulation or a scheduled recording with
auto-stimulate. See Chapter 4.

3.3.1. Setting Up for a Recording

1. Turn on the Middleman with the switch on the back.

2. Open AxIS.
3. Click the Environment Control module.
4. Turn on the heater to 37°C. If an Axion-compatible gas mixer is connected to the system, turn it on to
5% CO2. See Section 2.6.4 for more information about the Environment Control module.
5. Allow the system approximately 30 seconds to connect as indicated by the Maestro status light (See
Section 3.1 for more information on Status Lights).
Note: The data stream will say “Maestro” when connected and “No Device Connected” when it is
not or while initiating.
6. Allow the temperature and CO2 to stabilize.
7. Select the plate type from the Active Plate drop-down menu.
8. Double click on the Active Plate figure to enter a plate map.
Note: Plate maps may be imported from a .platemap file or a previously recorded .raw file.
9. Right-click on the Maestro or Muse stream and select Configuration  Cardiac Real-Time or Neural
Real-Time  Spontaneous, Electrically Evoked, or Optically Evoked to apply a configuration. See
Section 3.2 for more information on hardware configuration and software configurations.

10. Click the Experiment Setup Properties module. See Section 2.6.1 for more information on the
Experiment Setup Properties module.

[41]
AxIS 2.4 User Guide

11. Click Browse to set the save location in the Experiment Folder field.
12. Select AxIS Raw from the Maestro or Muse Continuous Stream drop-down to save the continuous
waveform data output. Leave all other streams set to Not Recording. See Section 5.2 for more
information on output file types.
13. Optional: Uncheck Auto Name File below the Maestro menu and enter a file name. By default, the
name will be Maestro_. Auto-naming macros are available to add descriptive information to file
names. See Section 5.2.1 for more information on naming output files.
Note: File names are appended with a number starting with 000. Recording a file with the same
name will increase the number appended to the end so no two files share a name.
14. Optional: Enter the Investigator, Experiment ID, and Description in the respective fields.

Set Save Location

Experiment Details

File Name Save AxIS Raw

Experiment Setup

3.3.2. Recording from the Maestro Manually

1. Open the ECMini lid.

2. Lift the Maestro handle, place the MEA plate into the docking bay and lower the handle to engage it
to the system.
Note: For a Maestro with an automated docking bay, place the docking bay into the Open position,
place the MEA plate into the docking bay, and press the button to engage the MEA.
3. Close the ECMini lid.
4. Ensure gas is flowing to the ECMini when the MEA plate is on the system.
5. Set the ECMini stand gauge to 0.5 liters per minute.
6. Allow the MEA to equilibrate with temperature and CO2 engaged while docked in the Maestro for 5-
10 minutes prior to recording.

[42]
 Data Acquisition

7. Click Play to view the raw data and begin the voltage offset, indicated on the gray status bar. The
voltage offset runs each time AxIS begins measuring data from a stopped state and should be run
every time an MEA is placed on the system.
8. Wait for the voltage offset to complete (~10s in neural modes and 30s in cardiac modes).
9. Click the Record button to begin recording.
10. Click the Record or Stop button to stop recording.
11. Turn off the temperature and gas controls prior to exiting AxIS.

3.3.3. Recording from the Maestro using the Scheduled Recorder

1. Repeat Steps 1-7 from the previous section.

2. Click on the Scheduled Recording Setup module. See Section 2.6.2.
3. Enter the desired file duration in the Record For field.
4. Enter when to begin recording in the Starting field.
5. Enter how many times to create a recording in the Execute field. For a single recording, select Once.
6. If creating more than one recording from a file, enter the time between the start of each recording in
the Record Every field.
7. Click Start Schedule to begin recording.
8. Click Stop Schedule to stop recording manually after initiating a scheduled recording.

9. Turn off the temperature and gas controls prior to exiting AxIS.

3.3.4. Recording from the Muse

1. Slide open the Muse, place the MEA into the docking bay and slide closed to engage it to the system.
2. If recording in an incubator, allow the MEA to equilibrate with temperature and CO2 while docked in
the Muse for 5-10 minutes prior to recording.
3. Repeat Steps 7-10 from Section 3.3.2 to record manually or Steps 2-8 from Section 3.3.3 to record
using the Scheduled Recorder.
4. Turn off the temperature prior to exiting AxIS.

[43]
AxIS 2.4 User Guide

CHAPTER 4. STIMULATION
Stimulation offers a level of control over cell cultures to evoke specific activity or improve consistency between
wells and plates.

The Stimulation Studio module is used to design both electrical and optical stimulation protocols. Stimulation
Studio provides a variety of drag and drop blocks that can be used to build custom stimulation waveforms.

The Stimulation Studio has five major components:

1. Stimulation Blocks: Predefined functions representing the basic elements of a stimulus protocol.
2. Stimulation Lanes: Displays the currently assigned stimulation blocks.
3. Electrode or LED Selector: Assigns the protocol from the selected stimulation lane to specific
electrodes or LEDs.
4. Compiled Pattern View: Displays a composite of all stimulation lanes in a time/intensity plot.
5. Global Repeat Settings: Specifies how many times, and how often stimulation lanes repeat. These
settings apply globally to all lanes.

Stimulation Blocks Compiled Pattern View

Global Repeat Settings

Stimulation Lanes

Electrode or LED
Selector

Stimulation Studio

[44]
 Stimulation

4.1. ELECTRICAL STIMULATION

Using electrical stimuli applies a voltage or current waveform directly to the cells on the stimulation electrode.
It may be used to stimulate cells in a specific region of a well in their native state, without needing to
biologically modify the cells in any way.

Because the MEA system records an electrical signal (voltage), electrical stimulation results in a stimulus
artifact. This artifact is caused by both charge build-up on the electrode and saturation within the amplifiers. As
a result, AxIS recording is not accurate during the stimulation. An example of a stimulation artifact is shown
below in blue. The artifact is an initial biphasic waveform, followed by a slow recovery to baseline. In this case
the stimulus triggered a depolarization of the cell culture, referred to as a “capture”.

AxIS uses proprietary techniques to minimize stimulation artifacts. Built-in Neural Stimulation and Cardiac
Pacing blocks are provided that rapidly adjust filter settings to manage the stimulation artifact and drive the
electrode voltage back to pre-stimulation levels as quickly as possible. The amplifier outputs are disabled
when a stimulus is active. AxIS also manages stimulation artifact through built-in configurations, Electrically
Paced for cardiac and Electrically Evoked for neural. Both real-time and offline configurations are available
for data acquisition and analysis, respectively. Both configurations feature the Artifact Eliminator data
processor, to minimize the stimulation artifact (Section 2.3.3).

For data acquisition with electrical stimulation:

1. Build the stimulation protocol according to Section 4.1.2.

2. Record data according to Section 3.3, selecting the Cardiac Real-Time Electrically Paced or Neural
Real-Time Electrically Evoked configuration.
3. Click the Trigger button to begin the stimulation protocol manually. Click again to stop.
Note: If Once is selected in the Global Repeat Settings, the stimulation protocol may be started
automatically by the Scheduled Recording Setup module. See Section 2.6.2.

For data analysis with electrical stimulation:

1. Analyze data according to Section 5.5, selecting the Cardiac Offline Electrically Paced or Neural
Offline Electrically Evoked configuration.
Note: See the ‘Neural Metric Tool’ for evoked activity analysis options for neural analysis (Section
B.2.4).

[45]
AxIS 2.4 User Guide

4.1.1. Electrical Stimulation Blocks

The electrical stimulation blocks are:

Block Description
Recommended stimulation block for cardiac applications. Applies a
Biphasic Stimulation pulse (Voltage Stimulation mode) with artifact
elimination routine optimized for cardiac stimulation. Set duration with
Stimulus Duration in Pulse Settings dialog. Forces the voltage output to
Cardiac Pacing Stimulation match Voltage as long as Max Current has not been reached. Use
Stimulation Paddles for Estim+ Classic MEA 48 plates and Microelectrode
for all other MEAs.

Recommended stimulation block for neural applications. Applies a

Biphaisic Stimulation pulse (Voltage Stimulation mode) with artifact
elimination routine optimized for neural stimulation. Set duration with
Neural Stimulation with Stimulus Duration in Pulse Settings dialog. Forces the voltage output to
Artifact Elimination match Voltage as long as Max Current has not been reached.

Applies a single phase pulse of a set current or voltage for a set duration.
Use Stimulus Duration to set the duration.
Current Stimulation: Forces the current output to match Current as long as
Max Voltage has not been reached.
Monophasic Stimulation Voltage Stimulation: Forces the voltage output to match Voltage as long as
Max Current has not been reached.
Applies a dual phase pulse of a set current or voltage for a set duration.
Use Stimulus Duration to set the durations.
Current Stimulation: Forces the current output to match Current as long as
Max Voltage has not been reached.
Biphasic Stimulation Voltage Stimulation: Forces the voltage output to match Voltage as long as
Max Current has not been reached.

Applies a wait period. Set the duration using the Duration field.

Delay

Discharges the electrode to pre-stimulus voltage. Length of discharge set by

Discharge Duration. The transition from the stimulating to non-stimulating
state is controlled by the time constant Soft Switch. Discharge Strength sets
Stimulation Artifact the maximum current used to "pull" charge off the electrode and to ground.
Elimination

[46]
 Stimulation

Repeats any blocks contained within a set number of times. Double-click

the repeat number (xN) in the top right corner of the block to set the
number of times.
Loop Container
Set the location of a “tag” or time stamped note to indicate a stimulation
occurred. Tags are useful for visualization and analysis of evoked activity.
Only one tag block can be set per stimulation protocol, so tags cannot be
assigned to multiple lanes of stimulation.
Electrical Stimulation Tag Note: If no tag is manually set, a tag will be automatically generated at
each repetition of Lane A. To prevent excessive size of output files, limit the
tag rate to 50 Hz.

4.1.2. Stimulation Protocol Generation

A stimulation protocol should have four goals:

1. Maximize the efficacy, or capture, of the stimulus.

2. Minimize the damage to the cells.
3. Minimize the damage to the electrodes.
4. Minimize the stimulation artifact.

Cell damage can result from high charge injection or charge density, while damage to the electrodes is
caused by electrolysis. For microelectrodes of this size, electrode damage is more likely than cell damage.
These risks and the stimulus artifact increases with the length of stimulation, but can be significantly reduced by
using a charge-balanced stimulus shape (biphasic) and the Stimulation Artifact Eliminator block. The
recommended Neural Stimulation block and Cardiac Pacing Stimulation block contain these features. In
general, choosing the smallest charge-balanced, biphasic, voltage-controlled pulse with an effective capture
rate is recommended.

(Refer to Wagenaar et al. for an in depth discussion of stimulus waveform shape: Wagenaar, DA, Pine, J, and
Potter, SM. Effective parameters for stimulation of dissociated cultures using multi-electrode arrays. J.
Neurosci. Methods. 138(1-2), 27-37 (2004))

To evoke activity from neuronal cultures, the Neural Stimulation Block is recommended. This block contains an
Electrical Stimulation Tag, a Biphasic Stimulation, and a Stimulation Artifact Eliminator optimized for neural
cultures. The most important parameter for modulating the stimulus efficacy is the Stimulus Duration. A typical
value is 0.2-4 ms for standard electrodes, or 0.2-1 ms if using the dedicated stimulation electrode on an E-
stim+ Classic MEA48 plate. Use a duration less than 0.5 ms for the best artifact elimination.

To evoke activity from cardiac cultures, the Cardiac Stimulation Block is recommended. This block contains an
Electrical Stimulation Tag, a Biphasic Stimulation, and a Stimulation Artifact Eliminator optimized for cardiac
cultures. The most important parameters for modulating the stimulus efficacy is the Stimulus Duration and the
number of stimulus electrodes. Using an E-Stim+ Classic MEA 48 plate, only the stimulation paddle is used
with a duration of 0.2-1 ms. For microelectrode stimulation, a duration of 1-4 ms is typically used with 2-4

[47]
AxIS 2.4 User Guide

adjacent electrodes. It is recommended to start with the shortest duration and fewest number of electrodes and
increase until consistent capture is achieved (See recommended workflow in Section 4.1.3).

Note: The Neural Stimulation and Cardiac Pacing Stimulation blocks already contain a stimulation time tag.
Do not place an Electrical Stimulation Tag block directly in front of these blocks, or the built-in artifact
elimination will not work.
It is recommended to always save a stimulation protocol for future reference after the pattern is completed and
the electrodes are selected. Save stimulation protocols using File  Save Stim Waveform. Open with File 
Open Stim Waveform.

To build an electrical stimulation pattern:

1. Select Electrical in the Stimulation Type drop-down menu.

2. Click and drag a block into the desired stimulation lane. A gray bar appears in the lane indicating
where a block will be dropped.
3. Double-click a block in a stimulation lane to change its settings.
Note: Use the shortest duration stimulus that is effective in order to minimize the stimulation artifact.

To copy a block from a stimulation lane:

1. Right-click on the block and select Copy Pulse.

2. Right-click on the stimulation lane and choose Paste Pulse.

To delete a block from a stimulation lane:

1. Right-click on the block and select Delete Pulse.

Note: A Loop Container provides the option of deleting the loop and its contents or just the loop.

To clear a stimulation lane:

1. Right-click on the lane and select Clear.

To specify how often to apply the stimulation:

1. Select an option from the Stimulate drop-down menu in the Global Repeat Settings bar

To assign a stimulation lane to one or more electrodes:

1. Select the plate type from the drop-down menu above the Electrode Selector in Stimulation Studio.
2. Click on the desired stimulation electrodes in the electrode selector to the right of the plate map.
Selected electrodes will display an “A” and are highlighted with a red square. Click All to assign all
electrodes, or Clear to un-assign all electrodes.
Note: E-Stim+ Classic MEA 48 plates automatically assign electrode 41 (the dedication stimulation
electrode) when selected.

[48]
 Stimulation

3. Click on the desired wells that should receive the stimulation pattern. Selection is the same as the Plate
Map Editor in Active Plate (See Section 2.2.3).
4. Click Apply. The stimulation electrodes in the electrode selector will be applied to the wells in the
plate map.
Note: To remove stimulation electrodes from a well, “apply” the electrode selector to that well with no
stimulation electrodes selected.

Plate Type

Selected Wells Selected Electrodes

Apply

4.1.3. Example Neural Stimulation Patterns

To build the recommended neural stimulation protocol:

1. Select Electrical in the Stimulation Type drop-down menu.

2. Click and drag the Neural Stimulation with Artifact Elimination block into Lane A.
3. Optional: Double-click the block to change its settings.
4. Select an option from the Stimulate drop-down menu in the Global Repeat Settings bar.
Note: The Neural Stimulation block already contains a stimulation time tag. Do not add an Electrical
Stimulation Tag block.

[49]
AxIS 2.4 User Guide

To build the custom neural stimulation protocol shown below:

1. Select Electrical in the Stimulation Type drop-down menu.

2. Click and drag an Electrical Stimulation Tag block into Lane A.
3. Click and drag a Loop Container into Lane A.
4. Double-click the x2 in the upper right corner of the Loop Container and change it to x5.
5. Click and drag the Biphasic Stimulation block into the Loop Container.
6. Click and drag the Stimulation Artifact Eliminator block into the Loop Container behind the Biphasic
Stimulation block.
7. Click and drag the Delay block into the Loop Container behind the Stimulation Artifact Eliminator
block.
8. Select Once from the Stimulate drop-down menu in the Global Repeat Settings bar.

Note: Custom electrical stimulation waveforms will have larger artifacts because the Artifact Eliminator only
works with the Neural Stimulation and the Cardiac Pacing Stimulation blocks in Stimulation Studio. Axion
recommends using the Cardiac Pacing Stimulation block to pace cardiomyocytes and the Neural Stimulation
block for all other electrical stimulation applications.

4.1.4. Cardiac Pacing Stimulus Design Workflow

A few quick test stimuli are recommended to determine the appropriate stimulus for reliably capturing the wells
of interest before beginning the full stimulation protocol.

For Microelectrode stimulation, the suggested workflow is shown below. Proceed until successful capture is
achieved at the desired maximum frequency. In general, the required stimulus depends on cell type, plating
density/uniformity, and culture viability. Choose the lowest level stimuli that achieves complete capture at the
highest pacing rate.

Note: It is not possible to pace slower than the spontaneous frequency of the cardiac culture. Be sure to
choose the pacing rate with both the control frequency and the predicted pharmaceutical effects in mind.
When using treatments that will prolong the beat period, choose a slow pacing rate; when shortening the beat
period, choose a faster pacing rate.

[50]
 Stimulation

Stimulating using the dedicated stimulation electrode (Stimulation Paddles) on the E-Stim+ MEA Plate is
generally more effective at capturing the network than stimulating through the microelectrodes. If the default
parameters are not sufficient to capture the culture, change the stimulation by:

1. Increasing the duration to 0.5 ms, and decreasing the amplitude to 20 µA.
2. If the stimulation still does not capture the culture, increase the duration and/or amplitude in small
increments (0.1 ms and/or 10 µA), not to exceed 1ms and 100 µA. Faster pacing frequencies may
require larger stimulation parameters than slower pacing frequencies.
3. If the stimulation still does not capture the culture, change the pacing frequency to a frequency closer to
the culture’s spontaneous beat rate.

As the duration and amplitude of stimulation are increased, stimulation artifact may begin to increase. It is
important to make sure the beats detected by AxIS are true beats and not stimulation artifact. Check the
Conduction Plot in the Cardiac Beat Plots module to make sure the beat propagation is in the physiological
range 2-10 ms. If the conduction plot shows propagation less than 1.5 ms along with an unusual pattern, the
Cardiac Beat Detector is most likely detecting simultaneous stimulation artifact rather than propagation across
the syncytium. Alternatively, use the Stimulation Inspector (Section 2.3.4) to determine the size of the
stimulation artifacts. Increase the Beat Detection Threshold to prevent detection of stimulation artifacts as beats.

True Beats Stimulation Artifacts

[51]
AxIS 2.4 User Guide

4.2. OPTICAL STIMULATION

Axion designed the Lumos™ optical stimulator for artifact-free light stimulation. Cell cultures must be
engineered with opsins in order to become light-sensitive. When integrated with AxIS, the Lumos can
independently stimulate each well with up to four wavelengths. The optical stimulation interface in Stimulation
Studio is laid out similarly to the electrical stimulation interface.

Stimulation Blocks Compiled Pattern View

Global Repeat Settings

Stimulation Lanes

LED Selector

For data acquisition with optical stimulation:

2. Build the stimulation protocol according to Section 4.2.2.

3. Record data according to Section 3.3, selecting the Cardiac Real-Time Optically Paced or Neural
Real-Time Optically Evoked configuration.
4. Click the Start button to begin the stimulation protocol manually. Click again to stop.
Note: If Once is selected in the Global Repeat Settings, the stimulation protocol may be started
automatically by the Scheduled Recording Setup module (Section 2.6.2).

For data analysis with electrical stimulation:

1. Analyze data according to Section 5.5, selecting the Cardiac Offline Optically Paced or Neural
Offline Optically Evoked configuration.
Note: See the ‘Neural Metric Tool’ for evoked activity analysis options for neural analysis (Section
B.2.4).

[52]
 Stimulation

4.2.1. Optical Stimulation Blocks

The optical stimulation blocks are:

Block Description

Applies a constant intensity of light for a set duration. Set intensity and
duration with Intensity and Stimulus Duration in the Pulse Settings dialog.
Optical Pulse Stimulation
(On)

Applies a set duration of no light. Use as a wait period. Set duration with
Stimulus Duration in Pulse Settings dialog.
Optical Pulse Stimulation
(Off)

Applies a linearly increasing intensity for the duration of the pulse. Set the
starting and ending intensity and duration with Starting Intensity, Ending
Intensity, and Ramp Duration in the Ramp Settings dialog.
Optical Ramp Stimulation

Repeats any blocks contained within a set number of times. Double-click

the repeat number (xN) in the top right corner of the block to set the
number of times.
Loop Container
Set the location of a “tag” or time stamped note to indicate a stimulation
occurred. Tags are useful for visualization and analysis of evoked activity.
Only one tag block can be set per stimulation protocol, so tags cannot be
assigned to multiple lanes of stimulation.
Optical Stimulation Tag Note: If no tag is manually set, a tag will be automatically generated at
each repetition of Lane A. To prevent excessive size of output files, limit the
tag rate to 50 Hz.

4.2.2. Stimulation Protocol Generation

Stimulation Studio allows the user to generate multiple “lanes” of optical stimulation patterns. Each stimulation
lane may be assigned only one wavelength but multiple stimulation lanes may be assigned the same
wavelength. Each well may be assigned any combination of wavelengths, but no two stimulation lanes of the
same wavelength may be assigned to the same well.

Warning: Running LEDs at high output intensity settings with high duty cycles for extended periods of time will
result in significant build-up of radiated heat. Take special care not to overheat objects lying directly
underneath the Lumos, such as a culture plate, the Maestro, or the Lumos stand. If elevated temperatures are

[53]
AxIS 2.4 User Guide

maintained, ‘AxIS’ will turn the Lumos off, but it’s possible to reach temperatures that could be damaging to
cell cultures before then.
It is recommended to always save a stimulation protocol for future reference after the pattern is completed and
the wells are selected. Save stimulation protocols using File  Save Stim Waveform. Open with File  Open
Stim Waveform.

To build an optical stimulation pattern:

1. Select Optical in the Stimulation Type drop-down menu.

2. Click and drag a block into the desired stimulation lane. A gray bar appears in the lane indicating
where a block will be dropped.
3. Double-click a block in a stimulation lane to change its settings.
4. Click on the desired wavelength for the lane.
5. (optional) Type a name for the lane in the space beside the lane letter.

To copy a block from a stimulation lane:

3. Right-click on the block and select Copy Pulse.

4. Right-click on the stimulation lane and choose Paste Pulse.

To delete a block from a stimulation lane:

2. Right-click on the block and select Delete Pulse.

Note: A Loop Container provides the option of deleting the loop and its contents or just the loop.

To add new stimulation lanes:

1. Click the Add Lane button.

To copy a stimulation lane:

1. Right-click on the lane and select Copy Lane.

2. Right-click on a lane and select Paste Lane to create an identical new lane.
3. Right-click on a lane and Select Paste Lane Contents to paste the contents of the copied lane to the
current lane.

To clear a stimulation lane:

2. Right-click on the lane and select Clear.

To delete a stimulation lane:

1. Click on the red X ( ) in the top right corner of the lane.

To hide a stimulation lane:

[54]
 Stimulation

1. Click on the eye icon ( ) in the bottom right corner of the stimulation lane.
Note: Hiding a lane will only prevent it from being seen in the compiled pattern view, not disable its
activity during stimulation.

To specify how often to apply the stimulation:

1. Select an option from the Stimulate drop-down menu in the Global Repeat Settings bar
Note: The Global Repeat Settings bar applies to all stimulation lanes defined in the stimulus. To set a
distinct frequency, or number of repetitions, across different lanes, use the repeat block in each individual
lane.

To assign a stimulation lane to one or more wells:

1. Click on the stimulation lane to select it.

2. Click on the desired well(s) in the plate map in Stimulation Studio module to apply the lane to that
well. Selection is the same as the Plate Map Editor in Active Plate (See Section 2.2.3).
Note: When selected the lane color will become active with the lane letter inside.

Multiple stimulation lanes of the same wavelength can be defined, but each well may only be assigned a
single lane for each wavelength. Multiple stimulation lanes, each having a unique color, can be applied to the
same collection of wells. Right-click on the LED Selector to quickly clear previous well assignments.

Tag Marker

Letters indicate Lane

20 Hz pulse train at 75%

10 Hz pulse train at 40%

Right-click Menu
Constant orange light at 20%

Constant green light at 5%

[55]
AxIS 2.4 User Guide

4.2.3. Example Stimulation Pattern

A basic example of a periodic blue light stimulus is shown below. The pulse block is set to 5 ms duration at
25% intensity. The lane is designated for the blue LED and is assigned to each well of the plate. The global
repeat bar specifies a 1 Hz frequency. This stimulation could be used to pace a spontaneously beating
cardiac culture with a periodic blue light stimulus.

Wavelength Select Well Select

Add Lane

Start stimulation

Use the following steps to recreate the example stimulus above:

1. Select Optical from the Stimulation Type drop-down menu.

2. Click and drag an Optical Stimulation Tag block into Lane A.
3. Click and drag an Optical Pulse Stimulation (On) block into Lane A.
4. Double-click the On block to set the Intensity to 25% and Stimulus Duration to 1 ms.
5. Click the blue button to the left of Lane A to specify the wavelength as 475 nm (blue).
6. Click on the * in the upper left corner of the plate map to assign the stimulus pattern in Lane A to all
wells in the plate.
7. Select until stopped in the Stimulate drop-down. Set every to 1 second(s).

4.3. EVALUATING CARDIAC STIMULATION CAPTURE

Effective capture means each stimulation pulse elicits a beat throughout each of the selected wells. This is the
most critical factor when determining stimulation parameters. If a stimulus does not capture the majority of
wells, a stronger stimulus should be used (see Section 4.1.2 for recommendations). Generally, the “weakest”
stimulus that consistently captures each well should be used when possible. Stronger stimuli can affect the
ability to record the sodium spike, due to the larger stimulation artifact. The Beat Periods plot in the Cardiac
Beat Plots module, the Continuous Waveform Plots module, and the Activity Map module are all useful in
diagnosing whether a well has been captured.

[56]
 Stimulation

View the Beat Periods plot to see if the beat rate matches the stimulus rate.

Spontaneous Beat Period Paced Beat Period

Capture is evident in the Continuous Waveform Plots module by looking for the depolarization spike and T-
wave for each beat. When a depolarization spike and T-wave follow each stimulation artifact, capture is
successful.

The figure below shows a partial capture in the Continuous Waveform Plots module. Most, but not all, stimuli
show a corresponding cardiac beat.

Failed Stimulus

No T-Wave
T-Wave

[57]
AxIS 2.4 User Guide

Capture can be assessed using the Activity Map with the source set to Beat Rate. Set the scale to the target
paced beat rate. When the beat rate of each electrode is equal to the pacing rate, the well has been
successfully paced.

The following table illustrates how various capture conditions appear in the Activity Map and Continuous
Waveform Plots modules:

Status Activity Map Continuous Waveform Plot

Full Capture
Evoked Beat
All stimuli elicit beats T-Wave

Stimulation Artifact

Partial Capture
Evoked Beat T-Wave
Some stimuli elicit beats

Stimulation Artifact No T-Wave

No Capture
No stimuli elicit beats

Stimulation Artifact

In order to use the Beat Periods plot and Activity Map to evaluate capture, ensure the Beat Detector is
capturing true beats, not stimulation artifact (Section 2.3.7).

[58]
 Stimulation

4.4. EVALUATING NEURAL STIMULATION CAPTURE

Neural stimulation capture can be viewed in the Spike Plots module with a combination of Electrical or Optical
Stimulation Tags and the Stimulation Inspector data processor.

The stimulation tags synchronize the recorded electrophysiological activity with the stimulation pattern. The
timing of each stimulation is displayed alongside the recorded electrophysiological data, providing instant
feedback on a successful stimulus. The tags are displayed in the raster plot as inverted white triangles,
corresponding to the tags defined in Stimulation Studio.

Tag

The Stimulation Inspector data processor plots the continuous voltage data from before and after an Electrical
or Optical Stimulation Tag to the Spike Plots module. The waveform plots overlay with the brightest trace
representing the most recent tag. This provides a close look at data near a Tag to evaluate artifact elimination
and cell response to the stimulation. See Section 2.3.4 for more information on the Stimulation Inspector.

[59]
AxIS 2.4 User Guide

CHAPTER 5. DATA ANALYSIS

From a functional standpoint, data analysis in AxIS is similar to data acquisition. A previously recorded .raw
file (continuous voltage data) is loaded into the Streams window, it passes through various data processors,
and output files are generated while AxIS “records”. Instead of recording the continuous voltage data from a
Maestro or Muse stream, file outputs from the data processors with various endpoint measurements are
generated. This chapter reviews how to analyze data from the Maestro/Muse using AxIS. Sections 5.5
and 5.6 contain step-by-step tutorials for analyzing a single file and multiple files, respectively.

5.1. STREAM CONFIGURATIONS

Similar to data acquisition, AxIS comes with a variety of preset configurations for data analysis. Use the
Offline configurations for data analysis.

To apply an analysis configuration:

1. Right-click on the data stream.

2. Select Configuration and navigate to the desired configuration. Click on the configuration.

Offline analysis configurations available in AxIS include:

Configuration Processing Applied Description

Cardiac Offline
Spontaneous Digital Filter Generates activity map, cardiac
Cardiac Beat Detector waveforms, beat rate and conduction
Statistics Compiler information. FPD detection is enabled.
Generates a variety of cardiac beating
endpoints.
Electrically Artifact Eliminator Generates activity map, cardiac
Paced Cardiac Beat Detector waveforms, beat rate and conduction
Statistics Compiler information. Optimized for reducing
stimulus artifacts in pacing experiments.
FPD detection is enabled. Generates a
variety of cardiac beating endpoints.

[60]
 Data Analysis

Optically Paced Digital Filter Generates activity map, cardiac

Cardiac Beat Detector waveforms, beat rate and conduction
Statistics Compiler information. FPD detection is enabled.
Generates a variety of cardiac beating
endpoints.
Neural Offline
Spontaneous Digital Filter Generates activity map, spike waveforms,
Spike Detector and raster plot. Network burst detection is
Burst Detector enabled. Generates a variety of spiking,
Statistics Compiler bursting, and synchrony endpoints.
Electrically Digital Filter Generates activity map, spike waveforms,
Evoked Artifact Eliminator and raster plot. Optimized for reducing
Spike Detector stimulus artifacts in electrically-evoked
Burst Detector experiments. Network burst detection is
Statistics Compiler enabled. Generates a variety of spiking,
bursting, and synchrony endpoints.
Optically Filter Generates activity map, spike waveforms,
Evoked Spike Detector and raster plot. Network burst detection is
Burst Detector enabled. Generates a variety of spiking,
Statistics Compiler bursting, and synchrony endpoints.

Custom analysis configurations can be designed by adding data processors to the data stream (Section
2.3.1). For instructions on saving and loading custom stream configurations, see Section 3.2.2.

5.2. OUTPUT FILE TYPES

The AxIS Raw (.raw) file is the most fundamental AxIS file type; it contains the continuous voltage data.
Any downstream processing can be recreated when the .raw file is loaded into AxIS. In addition to the
.raw file, AxIS can generate a variety of file outputs from the various data processors.

Data Processor Output Name Description

Continuous Streams
Maestro/Muse AxIS Raw (.raw) Primary format for data acquisition. Continuous voltage data
from all enabled electrodes before any digital processing.
Digital Filter AxIS Raw (.raw) Continuous voltage data from all enabled electrodes after
the application of a Digital Filter.
Artifact Eliminator AxIS Raw (.raw) Continuous voltage data from all enabled electrodes after
the application of an Artifact Eliminator.
Spikes
Spike Detector AxIS Spike (.spk) Spike voltage waveforms with time and electrode for all
spikes detected by the Spike Detector. Required output file
for Neural Metric Tool.

[61]
AxIS 2.4 User Guide

Spike Counts (.csv) Table listing the number of spikes detected in a given time
period for each electrode. Spike counting interval is set in the
Spike Detector Settings.
Spike List (.csv) Table listing time, electrode, and amplitude of all spikes
detected by the Spike Detector.
Burst Detector Electrode Burst List Table listing the individual electrode bursts identified by the
(.csv) Burst Detector with descriptive information including time,
electrode, number of spikes, and duration.
Network Burst List Table of the network bursts identified by the Burst Detector
(.csv) with descriptive information including time, well, number of
spikes, duration, number of electrodes, and spikes per
electrode.
Cardiac Beats
Cardiac Beat Electrode Beat List Table of individual electrode beats with time, electrode,
Detector (.csv) depolarization amplitude, depolarization slope, beat
period, field potential duration, well beat number, and delay
from well beat time.
Well Beat List (.csv) Table of well beats with well, time, number of electrodes,
starting electrode, ending electrode, beat period, well beat
number, conduction velocity, and maximum delay from well
beat time.
Advanced Statistics
Cardiac Statistics Advanced Metrics Primary output for data analysis. Tables containing group,
Compiler (.csv) well, and electrode endpoint metrics averaged over the
duration of the analysis window. Includes depolarization
amplitude, slope, beat period, beating irregularity, field
potential duration, and conduction velocity metrics. Required
output file for CiPA Analysis Tool and/or AxIS Metric
Plotting Tool. See Section 6.3 for a list of included endpoints.
Neural Statistics Advanced Metrics Primary output for data analysis. Tables containing group,
Compiler (.csv) well, and electrode endpoint metrics averaged over the
duration of the analysis window. Includes spike, burst, and
synchrony metrics. Required output file for AxIS Metric
Plotting Tool. See Section 7.4 for a list of included endpoints.

The Experiment Setup Properties module defines the output file types that will be saved. Select the desired
output from the drop-down menus in each section. More than one output may be selected from any
drop-down menu. Any stream that should not be recorded must be set to Not Recording. Axion recommends
saving an AxIS Raw file from the Maestro/Muse stream during data acquisition. Generally, it is not necessary
to save any other AxIS Raw files during acquisition nor analysis. Preset Real-Time and Offline stream
configurations specify the recommended output files to save.

[62]
 Data Analysis

Inactive Stream

Select File Type(s)

Add Output File

To select multiple outputs from a single drop-down in Experiment Setup Properties:

1. Select the first output from the drop-down.

2. Click the Add button ( ).
3. Select the next output.
4. Repeat Steps 1-3 as needed.
Note: To remove an output, click the Remove button ( ) or select Not Recording from the drop-
down menu.

AxIS supports the use of third party software, including NeuroExplorer®, Microsoft Excel®, Offline
Sorter®, MATLAB®, Spotfire®, and any software that can import comma-separated value (.csv) data. To
use these software packages, AxIS data must be recorded in the appropriate file types. These files will
contain different information depending on the application. AxIS Raw files can be processed in AxIS,
MATLAB, or NeuroExplorer. AxIS Spike files can be processed in MATLAB, NeuroExplorer, and Offline
Sorter. Spike List, Spike Count, Burst List, Beat List, and Advanced Metrics files can be processed in
Excel.

5.2.1. Recorded File Names

AxIS automatically names files based on the continuous data stream and the data processor used to
generate the output. AxIS will append a 3-digit number to the end of every file name (manually or
automatically named) to prevent identical file name conflicts. The number starts at 000 and increases by
1 for each file recorded to the same location with the same file name. As an example: an Advanced

[63]
AxIS 2.4 User Guide

Metrics file generated from a .raw file named “Compound A.raw” will be named “Compound
A_Statistics Compiler_(000).csv”.

Experiment properties such as date, duration, or stream source can be auto-generated in the file name
using macros. A file name may be a mixture of macros and user entered text. The previous example of
“Compound A_Statistics Compiler_000” could be manually entered with macros as
[SourceFile]_[StreamName].

Available macros are listed in the table below:

Macro Description Example

File path (example, Places the output files in the specified folder instead Places file in the designated
C:\) of the source file folder. Available for batch folder
processing only.

[Date] Date the output file is created. (yyyymmdd) 20161210

[Duration] Length of the output file. 1m0s

[ExperimentID] Experiment ID as entered in the Experiment Setup Experiment1

Properties module.

[FileSegment] Portion of the file was analyzed. Start of File

[Investigator] Investigator as entered in the Experiment Setup SC

Properties module.

[Offset] Time from the start of the file processing began. 10s

[SourceFile] Original filename of the source .raw file. Compound_A

[StreamName] Name of the data processor generating the output Statistics Compiler
file.

To manually name a file:

1. Deselect the Auto-Name File checkbox in the Experiment Setup Properties module.

To add a macro to the file name:

1. Type the macro command from the above table into the Filename field.
– Or –
1. Right-click the Filename field.
2. Select Macros and choose the macro from the menu.

[64]
 Data Analysis

Auto-name File

File Name Macros

Enter File Name

The Filename field can create folders that do not currently exist. Use ‘\’ to create new paths for output
files. For example, the entry [SourceFile]\[StreamName] for the previous example will name the output
file “Statistics Compiler_000” and place it in a new folder called “Compound A”, created in the folder
containing the “Compound A.raw” file.

5.3. PREPARING FILES FOR ANALYSIS

Before beginning an analysis ensure all files are appropriately named and contain accurate plate maps,
notes, experiment IDs, and descriptions. These notes will be kept in the analysis outputs. All output files are
automatically stored in the same folder as the .raw files so organize the .raw files in appropriate folders before
beginning.

5.4. SELECTING THE ANALYSIS WINDOW

While performing an analysis, AxIS will stream through the continuous voltage data. AxIS will process the
entire file or a segment of the file specified in the Stream Settings. A portion of the file can be analyzed
manually by starting and stopping the data stream and recording the region of interest using the Play, Pause,
Stop, and Record buttons in the Display Controls (See Section 2.4).

To set the analysis window automatically:

[65]
AxIS 2.4 User Guide

1. Click Stop if AxIS is not already in a stopped state.

2. Right-click the file in the Streams window.
3. Select Settings.
4. Set the analysis window using the Segment Type dialog.
Segment Type Description
Whole File Analysis window is the entire duration of the file.

Start of File Analysis window begins at the start of the recording

plus the Offset and continues for the Duration.

End of File Analysis window begins at the end of the file minus
the Duration and continues until the end.

5. Click OK.

Check the Loop Playback box to replay the file from the beginning when the end has been reached. This is
useful when examining short files, but is not recommend while recording analysis files.

5.4.1. Re-recording Segments of .raw Files

AxIS provides a single measurement for each metric that is a mean over the entire analysis duration. To split a
raw file into segments to obtain a time course, for example, or to archive only a portion of the file, it is possible
to record shorter .raw files from a .raw file.

1. Click File  Open Recording…

2. Select the file for analysis and click Open.
3. Set the analysis window. See Section 5.4.
4. Click the Experiment Setup Properties module.
5. Select AxIS Raw from the Maestro or Muse drop-down.
6. Optional: Uncheck Auto Name File to manually enter a file name. By default, the name will be
[Date][SourceFile]. See Section 5.2.1 for more information on naming output files.
7. Click Record.

To generate a time course from a single .raw file:

Select an analysis window and run an analysis according to Section 5.5. Repeat for each analysis
window.
– Or –
Generate new .raw files for each analysis window and run a batch process according to Section 5.6.

[66]
 Data Analysis

5.5. DATA ANALYSIS TUTORIAL

1. Click File  Open Recording…

2. Select the file for analysis and click Open.
3. Prepare the file for analysis. See Section 5.3.
4. Set the analysis window. See Section 5.4.
5. Right-click on the file in the Streams window and select Configuration  Cardiac Offline or
Neural Offline  Spontaneous, Electrically Paced/Evoked, or Optically Paced/Evoked to
apply a configuration. See Section 5.1 for software configurations.

6. Click the Experiment Setup Properties module. See Section 2.6.1 for more information on the
Experiment Setup Properties module.
7. Select Advanced Metrics from the Statistics Compiler drop-down. Select any additional desired
file outputs. It is not necessary to record the .raw data again. See Section 5.2 for more
information on output file types.
8. Optional: Uncheck Auto Name File beside the selected file outputs to manually enter a file name.
By default, the name will be [SourceFile]_[StreamName]. See Section 5.2.1 for more information
on naming output files.
9. Click Play to view the data.
10. Assess the analysis settings. See Chapter 6 for cardiac activity, Chapter 7 for neural activity.
Note: If the file is a short recording, right-click the file in the Streams window and select Settings. Click
Loop Playback to enable and click OK. This automatically replays the file when it reaches the end,
providing more time to assess analysis settings. Make sure to disable Loop Playback prior to
recording analysis files.
11. Right-click on the file in the Streams window and select Settings.
12. Click Accelerate Playback to enable. Accelerate Playback plays the file and performs analysis
faster than real time.
13. Click OK.
14. Click Record to run the analysis.
Note: If setting the analysis window manually, start and stop the recording at the desired times.

[67]
AxIS 2.4 User Guide

5.6. ANALYSIS OF MULTIPLE FILES (BATCH PROCESSING)

It is possible to analyze multiple files sequentially using identical analysis settings with batch processing.
In this mode analysis can be done without having to queue each file individually and wait for the
processing to finish.

A batch process appears as a stream in the Streams window. The Batch Process stream acts like a
traditional stream; analysis configurations may be applied to the batch process in the same manner as a
single file, and all files in the batch process will be processed according to the stream configuration.
When a batch process is the active stream, the Play button becomes a Start Batch Process button.

Start Batch
Selected Files

Output Files

Batch Process Add Files

Segment Type

To begin a batch process:

1. Click File  New Batch Process...

2. Click Add in the Edit Batch Process Settings dialog.
Note: Use Remove to remove any files from the Source File list.
3. Select the desired .raw files and click Open.
4. Select the analysis window using the Segment Type drop-down menu. See Section 5.4 for setting
an analysis window with the Segment Type dialog.
Note: An orange file indicates the segment is longer than the total length of the file. Analysis will
stop when the file end is reached. If the file is so short that nothing will be analyzed, the file entry
turns red.
5. Click OK.
6. Right-click on the batch process in the Streams window and select Configuration  Cardiac
Offline or Neural Offline  Spontaneous, Electrically Evoked, or Optical Evoked to apply a
configuration. See Section 5.1 for more information about analysis configurations.

[68]
 Data Analysis

7. Click the Experiment Setup Properties module. See Section 2.6.1 for more information on the
Experiment Setup Properties module.
8. Select Advanced Metrics from the Statistics Compiler drop-down. Select any additional desired
file outputs. It is not necessary to record the .raw data again. See Section 5.2 for more
information on output file types.
Note: File names are automatically generated as [SourceFile]_[StreamName] appended with a
number starting with 000, and output files are saved to the same directory as the source .raw file
each was generated from.
9. Click Start Batch Process.

A white status bar will appear at the top of the active window while a batch process is running. The status
bar will list the progress of the current file (File) and the entire batch process (Overall).

[69]
AxIS 2.4 User Guide

CHAPTER 6. CARDIAC DATA ANALYSIS

6.1. CARDIAC ACTIVITY INTRODUCTION

A cardiac action potential, or “beat”, consists of three phases, the depolarization, plateau, and
repolarization. These phases are governed by the balance of ionic currents during each phase. Measured
extracellularly on an MEA, these beats exhibit features similar to those present in an electrocardiogram
(ECG).

The initiation of a beat starts with the depolarization phase which corresponds to the QRS complex in the ECG
and the depolarization spike in the MEA recording. The T-wave in the ECG and the corresponding peak in the
MEA recording represent repolarization, when the voltage returns to baseline. For this document, the
repolarization peak in an MEA will be referred to as the T-wave. Not present in an MEA recording is a peak
corresponding to the P-wave. In an ECG the P-wave represents the depolarization of the atria while the QRS
complex and T-wave represent the depolarization and repolarization of the ventricles. A cardiac culture
typically lacks the structural compartmentalization of an intact heart and only has a single depolarization and
repolarization phase.

Under basal conditions a typical cardiac culture should beat rhythmically, with each cardiac waveform
identical to the last. From the cardiac beat waveforms numerous metrics can be evaluated to assess beat
rhythmicity and waveform stability. The time between depolarization spikes constitutes the beat period and
can be used to assess the beat rate and rhythmicity.

The time between the depolarization and repolarization is known as the field potential duration (FPD) and
corresponds to the QT interval in an ECG. Disrupting the balance of ion channels governing the plateau and
repolarization phases can alter the FPD. Prolongation in the FPD is assessed as a potential indicator of pro-
arrhythmia.

[70]
 Cardiac Data Analysis

The two-dimensional structure of the MEA allows cardiac signal propagation to be measured and quantified
by comparing the depolarization spike timing between electrodes. In culture, cardiomyocytes form an
electrically coupled syncytium of cells that beat synchronously. When a cardiomyocyte in that syncytium
depolarizes, it causes neighboring cardiomyocytes to depolarize, cascading throughout the culture. In vitro
cardiac beat conduction is analyzed by tracking the beat time across the array of electrodes, and calculating
the delay from where the beat originated.

AxIS will classify well beats by common starting and ending locations. In 1- and 12- well plates (64
electrodes per well), a loosely defined nearest neighbors grid of electrodes is considered a common starting
point; the most frequent starting electrode and any neighboring electrodes (vertically, horizontally, and
diagonally) are considered a common starting location. For 48-well plates, a stricter nearest neighbors
approach is used to define a common starting point; the most common starting electrode and neighboring
electrodes vertically and horizontally are considered a common starting location. For 96-well plates, only
beats beginning and ending at the exact same electrode are considered common. Each beat is assigned a
Propagation Classification ID, wherein 1 represents the most common propagation pattern detected in the
well, 2 is the next most common, and so on.

The figures below show four examples of propagation. The beat begins where the plot is blue (shortest delay
from the beat origin) and ends where the plot is red (longest delay from the beat origin). The stability of beat
propagation patterns and conduction velocity can be used to assess culture health, pacemaker stability, and
evaluate compound effects. Slowing or disrupting beat propagation can lead to arrhythmia in vivo.

[71]
AxIS 2.4 User Guide

6.2. IDENTIFYING CARDIAC DEPOLARIZATION AND REPOLARIZATION

6.2.1. Cardiac Beat Detection

Cardiac beat detection is based on the depolarization spike. The Detection Threshold of the Cardiac Beat
Detector data processor, should be set lower than the depolarization spike but greater than any other
waveform feature. Since AxIS calculates endpoints based on well beats, the detection threshold only needs to
be accurate on some electrodes, as set by Min Active Electrodes in the Cardiac Beat Detector. If AxIS
identifies an arrhythmia, ensure the threshold is properly set so only the depolarization spike, and every
depolarization spike, crosses the threshold. The threshold can be viewed from the Continuous Waveform Plots
module as horizontal gray lines above and below the data trace for each electrode.

Detection Threshold

6.2.2. T-Wave Identification

The T-wave is often much less pronounced than the depolarization spike and can vary considerably in size,
shape, and polarity from one electrode to the next. Use the Cardiac Beat Plots module to assess T-wave
detection accuracy as indicated by the white whisker plot.

Characteristics of a cardiac waveform from a synchronous cardiac culture can be represented by a subset of
the electrodes in each well, making it possible to eliminate electrodes containing noisy or anomalous data.
This is particularly useful for FPD measurements. T-wave locations, and therefore FPD measurements, should
be similar between electrodes; however, physical culture differences across the well can cause some
electrodes to display prominent T-waves and others to display none at all. For electrodes where the T-wave is
small or absent, AxIS may generate FPD values showing high beat-to-beat variability because of poor
detection performance.

[72]
 Cardiac Data Analysis

Waveform Description
Good
The T-wave has been properly and consistently Detection
detected. It shows a very small standard
deviation.

No T-wave
The trace shows no visible T- wave, therefore Present
detection is unreliable, resulting in a large
standard deviation.

Missed
The detection algorithm is alternately picking the
Detection
correct T-wave and a feature earlier in the beat.
The mean reflects the average of these two
humps and the standard deviation is very large.

To optimize T-wave detection:

1. Adjust the Post Spike Detection Holdoff, Pre Spike Detection Holdoff, and Max Post Search Duration
values. This sets the search range. If multiple peaks are present, AxIS will preferentially mark the first
peak. Adjust the search range to bracket the preferred peak.
2. Adjust the T-Wave Detection Feature to match the majority of T-waves. Max for upward deflecting
peaks, Min for downward deflecting T-waves.
3. Set the FPD Measure Quality Control settings to remove poorly detected electrodes. Arrhythmic wells
will have low beat to beat consistency and electrode FPD consistency.
4. Optional: If there are many electrodes that identify the wrong peak but have a high beat to beat
consistency, as indicated by the whisker plot, disable those electrodes.

Note AxIS will assess all electrodes in a plate by the same criteria. When analyzing plates with significantly
different waveforms between wells, it may be necessary to perform the analysis multiple times with settings
optimized for subsets of wells with similar waveforms. Use the CiPA Analysis Tool (Appendix C) to analyze
plates with highly variable or hard to analyze T-waves.

6.3. CARDIAC STATISTICS COMPILER ENDPOINTS

AxIS identifies depolarization spikes and T-waves with the Cardiac Beat Detector data processor (See
Chapter 6 and Section 2.3.7). It then calculates beat rate, waveform, and conduction metrics with the Cardiac
Statistics Compiler module and outputs an Advanced Metrics file with the results (See Section 2.3.8). The
Advanced Metrics endpoints are listed in the table below.

[73]
AxIS 2.4 User Guide

Measurement Description
Starting Time The start time of the analysis window. If Limit to Region of Most Stable
Beat Period is enabled, the start time of the most stable beat period.

Ending Time The end time of the analysis window. If Limit to Region of Most Stable
Beat Period is enabled, the end time of the most stable beat period.

Number of Beats Total number of well beats over the duration of the analysis.

Starting Electrode Electrode detected first during a well beat. The starting electrode of the
most common propagation pattern is listed in Well Averages.

Ending Electrode Electrode detected last during a well beat. The ending electrode of the
most common propagation pattern is listed in Well Averages.

Propagation Consistency The number of well beats that had the most common propagation
pattern divided by the total number of well beats.

Total Active Electrodes The number of electrodes that detected beats.

Total FPD Electrodes The number of electrodes remaining after FPD quality control
parameters from the Cardiac Statistics Compiler are applied.

Beat Period The time between successive depolarization spikes, in seconds.

Beat Period Irregularity The coefficient of variation (standard deviation/mean) of the beat
period multiplied by 100.

Spike Slope The maximum change in voltage over time (dV/dt) of the
depolarization spike, in V/s.

Spike Amplitude The peak to peak (positive plus negative) amplitude of the
depolarization spike, in mV.

FPD The time from the depolarization spike to the peak of the T-wave, in ms.

FPD COV The coefficient of variation (standard deviation/mean) of the FPD

multiplied by 100.

Conduction Velocity Speed the depolarization spike propagates across the culture,
calculated as an average for the entire well. The average is taken by
plotting the propagation delay of each electrode against its distance
from the beat origin. A best fit line created from these delays, and the
reciprocal of its slope is taken to obtain the velocity.

Max Delay The time between depolarization spike detection at the starting and
ending electrode.

[74]
 Neural Data Analysis

CHAPTER 7. NEURAL DATA ANALYSIS

7.1. NEURAL ACTIVITY INTRODUCTION

Neural cultures form intricate networks of cells that produce complex patterns of activity. A neural action
potential consists of depolarization and repolarization phases governed by the balance of ionic currents
during each phase. Examples of neuronal action potentials measured intracellularly and extracellularly are
shown in the figure below.

Intracellular Action Potential Extracellular Action Potential

Gold, et al., 2006. On the Origin of the Extracellular Action Potential

Waveform: A Modeling Study. J Neurophysiol, 95:3113-3128.

The first step to quantifying neural activity using an MEA is to identify individual neuronal action potentials or
“spikes”. The spike shape depends upon the cells’ proximity to the electrode. Neuronal activity on an MEA is
typically studied by quantifying the spike timing and how coordinated the spikes are across the culture.

7.2. IDENTIFYING NEURAL SPIKES

It is important to ensure the spikes detected are physiological. In general, physiological spikes will show a
consistent shape that happens repeatedly, while electrical noise will not. A single electrode however will
measure activity from all nearby neurons so more than one consistent spike waveform may be present. Spikes
can have non-standard shapes based on a neuron’s position with respect to the recording ground. Spikes
from multiple cells can sum together to create multiple peaks.

A quick check of the duration and amplitude of a spike can be helpful. Mammalian neurons produce action
potential widths of approximately 1-2 ms. Any signals significantly shorter should be considered carefully.
Amplitudes can vary greatly, but peak amplitudes typically range from 20 µV and 150 µV. Very large
amplitudes should be reviewed skeptically.

The following figure provides a series of examples of physiological spikes and electrical noise. These spikes
are easiest to evaluate within the Spike Plots module (Section 2.6.8).

[75]
AxIS 2.4 User Guide

Waveform Description Physiological/Noise

Consistent, repetitive
Physiological
monophasic waveform.

Consistent, repetitive
Physiological
biphasic waveform.

Multiple waveforms, but Physiological. Different neurons

each waveform is consistent. detected by the same electrode.

Physiological. Same neuron,

detection is near threshold
Waveforms are the same
causing different times in the
shape but out of phase.
waveform to be marked as the
start.

Single occurrence. Noise.

The same shape in both

positive and negative
directions. Unstable voltage Noise.
baseline before and after the
spike.

[76]
 Neural Data Analysis

7.3. TYPES OF NEURONAL ACTIVITY

Beginning with continuous voltage data (top) from a single electrode, AxIS identifies neuronal action
potentials, or “spikes”. Plotting just the spike time and location will generate a raster plot (below).
100 V

10 s

10 s

Neuronal activity on a single electrode may be quantified in a variety of ways. Firing may be random or
rhythmic, fast or slow, occur as single spikes or clusters. A cluster of spikes is called a burst. AxIS refers to
bursting on a single electrode as single-electrode bursting. From a burst, additional parameters may be
quantified such as burst frequency and duration.

Having multiple electrodes in an array recording allows examination of network-wide coordinated activity. A
network burst is a coordinated cluster of spiking across multiple electrodes. The distance between two
electrodes is such that no two electrodes will detect firing from the same neuron. Coordinated activity such as
a network burst therefore represents synaptic communication between neurons.

[77]
AxIS 2.4 User Guide

The level of coordinated or simultaneous spiking between electrodes is referred to as the culture’s synchrony.
Synchrony between two electrodes may be quantified using a cross-correlogram. The cross-correlogram
assesses the probability of a spike occurring on electrode A at times relative to a spike on electrode B. This
probability is summed across all spikes in electrode B to produce the cross-correlogram. For example, if both
electrodes always fire together, the cross-correlogram would have a sharp peak at time 0.

Measuring the area under the cross-correlogram around zero is an effective way to quantify synchrony. A
short synchrony window (e.g. 5 ms) quantifies synchrony on a millisecond timescale, while a long synchrony
window (e.g. 50 ms) captures synchronous activity on slower timescales.

AxIS uses frequency domain methods (Halliday, Rosenber, Breeze & Conway 2006) to compute and pool
the cross-correlogram across all unique pair-wise combinations of electrodes in a well. A normalized pooled
cross-correlogram is generated by removing remove autocorrelations, each electrode’s cross-correlation with
itself. An example of a pooled cross-correlogram is given below. When spiking occurs at similar times
between all electrodes, synchrony is high.

[78]
 Neural Data Analysis

7.4. NEURAL STATISTICS COMPILER ENDPOINTS

AxIS identifies spikes and bursting with the Spike Detector and Burst Detector data processors (See Sections
2.3.5 and 2.3.6). It then calculates spiking, bursting, and synchrony metrics with the Neural Statistics
Compiler module and outputs an Advanced Metrics file with the results (See Section 2.3.9). The activity can
be broadly grouped into four categories: spiking, single-electrode bursting, network bursting, and synchrony.
The Advanced Metrics endpoints are listed in the table below.

Measurement Description
Spiking
Number of Spikes Total number of spikes over the duration of the analysis.
Mean Firing Rate Total number of spikes divided by the duration of the analysis, in Hz.
ISI Coefficient of Variation The coefficient of variation (standard deviation/mean) of the inter-
spike interval, the time between spikes. This is a measure of spike
regularity.
Number of Active Electrodes Number of electrodes with activity greater than the minimum spike rate
set in the Neural Statistics Compiler.
Weighted Mean Firing Rate The mean firing rate based on only electrodes with activity greater than
minimum spike rate set by the Neural Statistics Compiler.
Single-Electrode Bursting
Number of Bursts Total number of single-electrode bursts over the duration of the
analysis.
Number of Bursting Electrodes Total number of electrodes in the well with single-electrode bursts.
Burst Duration Average time from the first spike to last spike in a single-electrode burst.
Number of Spikes per Burst Average number of spikes in a single-electrode burst.
Mean ISI within Burst Average inter-spike interval, time between spikes, for spikes in a single-
electrode burst. This is a measure of burst intensity; smaller values mean
more intense bursts.
Median ISI within Burst Median inter-spike interval, time between spikes, for spikes in a single-
electrode burst.
Inter-Burst Interval Average time between the start of single-electrode bursts.
Burst Frequency Total number of single-electrode bursts divided by the duration of the
analysis, in Hz.
Normalized Duration IQR Interquartile range of single-electrode burst durations. This metric
provides a measure of single-electrode burst duration regularity. If the
middle 50% of single-electrode bursts are approximately the same
duration, this value will be small, whereas, if the single-electrode bursts
vary widely in duration, this range will be large.
IBI Coefficient of Variation The coefficient of variation (standard deviation/mean) of the inter-burst
interval, the time between single-electrode bursts. This is a measure of
single-electrode burst regularity.

[79]
AxIS 2.4 User Guide

Burst Percentage The number of spikes in single-electrode bursts divided by the total
number of spikes, multiplied by 100.
Network Bursting
Number of Network Bursts Total number of network bursts over the duration of the analysis.
Network Burst Frequency Total number of network bursts divided by the duration of the analysis,
in Hz.
Network Burst Duration Average time from the first spike to last spike in a network burst.
Number of Spikes per Network Burst Average number of spikes in a network burst.
Number of Elecs Participating in Burst Average number of electrodes with activity during a network burst.
Number of Spikes per Network Burst Average number of spikes per burst divided by the number of
per Channel electrodes participating in that burst.
Network Burst Percentage The number of spikes in network bursts divided by the total number of
spikes, multiplied by 100.
Network IBI Coefficient of Variation The coefficient of variation (standard deviation/average) for the inter-
network burst interval, the time between network bursts. This is a
measure of network burst regularity.
Network Normalized Duration IQR Interquartile range of network burst durations. This metric provides a
measure of network burst duration regularity. If the middle 50% of
network bursts are approximately the same duration, this value will be
small, whereas, if the network bursts vary widely in duration, this range
will be large.
Synchrony
Area Under Normalized Cross- Area under the well-wide pooled inter-electrode cross-correlation
Correlation normalized to the auto-correlations. Higher areas indicate greater
synchrony.

Area Under Cross-Correlation Area under the well-wide pooled inter-electrode cross-correlation.
Full Width at Half Height of Cross- Distance along the x-axis (phase lag) from left half height to right half
Correlation height (probability) of the cross-correlogram. Higher full widths indicate
a wider correlogram (less synchrony) whereas lower full widths
indicate a taller correlogram (greater synchrony).
Synchrony Index A unitless measure of synchrony between 0 and 1 (Paiva et al 2010).
Values closer to 1 indicate higher synchrony.

[80]
 Axion Standalone Tools

APPENDIX A. AXION STANDALONE TOOLS

Axion provides a number of standalone tools that enable data post-processing and visualization. Providing
these tools independently AxIS allows Axion to rapidly respond to customer requests and provide additional
features that are not within the scope of the existing AxIS software. These tools use the MATLAB Compiler
Runtime package (MCR), which is automatically downloaded during the installation process.

The files required to install the Axion Standalone tools can be downloaded from Axion’s download site at
https://axionbiosystems.sharefile.com. Access to this site requires a log in and password. This should have
been provided when the Maestro/Muse was purchased. For account information, email
[email protected]. Log in to the download site and navigate to the Software/Tools folder to find the
standalone tool installation files.

Download the executable installation file for the desired tool. Double-click this file to begin installation and
follow the onscreen instructions. Most standalone tools work using the MATLAB Compiler Runtime program.
The installer will detect if the correct version of MCR is present on the computer and will download and install
the correct version, as needed. When the installation is complete, the standalone tool will appear in the start
menu and, if selected during installation, a shortcut will appear on the desktop.

To install the tool on a computer that does not have internet access, contact [email protected].

As Axion continues to develop new standalone tools to meet novel analysis needs, the functionality of older
standalone tools gets integrated in to the AxIS. As such, three standalone tools will no longer be supported
after AxIS 2.4:

Conduction Velocity Tool: Cardiac beat propagation maps and conduction velocity endpoints are now
available in the AxIS software (Sections 2.3.7 and 2.6.9). Users are encouraged to use AxIS for beat
propagation/conduction velocity analyses.

Cardiac Data Aggregation Tool: AxIS contains built-in beat quality control as well as FPD quality control
(Section 2.3.8), ensuring highly accurate reports of beating metrics. Users are encouraged to use AxIS for
automatic FPD analysis, and the CiPA Analysis Tool for applications requiring semi-automated oversight of
FPD measurements.

Cardiac Data Plotting Tool: The CiPA Analysis Tool provides beat waveform plots overlaid across dosing
conditions that be exported for each electrode and saved in a variety of file types (Appendix C). The beat
waveforms displayed in the Cardiac Beat Plots module in AxIS can be exported as either an image or the
source data points used to generate the image (Section 2.6.9). Users are encouraged to use the CiPA
Analysis Tool or the Copy Waveform feature in AxIS for cardiac data plotting.

[81]
AxIS 2.4 User Guide

APPENDIX B. NEURAL METRIC TOOL

B.1. INTRODUCTION
The Neural Metric Tool calculates bursting and synchrony metrics, similar to those computed by AxIS, along
with additional metrics. It also offers advanced algorithms for burst detection, stimulation-evoked activity
analysis, and generates plate map visualizations, raster plots, and synchrony cross-correlograms not
available in AxIS. See Chapter 7 for a detailed description of spiking, bursting, synchrony, and network
bursting.

B.2. NEURAL METRIC TOOL OVERVIEW

The Neural Metric Tool is divided into four sections: Plate Map, Bursting, Synchrony, and Evoked with four
corresponding analysis settings sections: Analysis Parameters, Bursting Parameters, Synchrony Parameters,
and Evoked Parameters. The Plate Map section displays results for the entire plate while Bursting, Synchrony,
and Evoked sections display data only for the active well (highlighted by the green box). Adjusting the settings
in a section and then clicking Apply will update the display sections above. Adjusting the settings in the
Analysis Parameters section will recalculate all spiking, bursting, synchrony, and evoked activity endpoints
and update all four display sections.

B.2.1. Plate Map Display and Analysis Parameters

The Plate Map displays plate-wide trends and between well comparisons for the selected neural metric; well
shading indicates the value of the metric chosen in the dropdown menu located above the Plate Map display.

The active well displayed in the Burst, Synchrony, and Evoked sections is selected by clicking on the Plate
Map. Use the arrow keys to navigate through the wells.

[82]
 Neural Metric Tool

To manually adjust the scale:

1. Deselect the Auto-scale checkbox

2. Enter high and low values in the top and bottom fields, respectively to the right of the scale
3. Click Refresh.

Displayed Metric Copy/Refresh & Autoscale

Active Well

The Analysis Parameters section defines the analysis window, activity criteria, and removes coincidence
artifacts. The analysis window is the file segment used to generate endpoint metrics. The activity criterion sets
the threshold for electrode inclusion in the weighted mean firing rate. Coincident artifacts are artificial spikes
detected on multiple electrodes at exactly the same time. They commonly occur during an electrical
stimulus or some environmental interference like bumping the system or touching the media in a well.

To set the analysis window

1. Select the segment type from the File Segment drop-down:

Segment Type Description
Whole File Analysis window is the entire duration of the file.
Start of File Analysis window begins at the start of the recording
plus the Offset and continues for the Duration.
End of File Analysis window begins at the end of the file minus
the Duration and continues until the end.
2. Click the Apply button in the Analysis Parameters section.

To specify the activity criterion:

1. Type a minimum firing rate (spikes/min) in the Active Electrode Criterion field.
2. Click the Apply button in the Analysis Parameters section.

[83]
AxIS 2.4 User Guide

To remove coincident artifacts:

1. Click the Remove Coincident Artifacts checkbox to enable. To blank spikes after a coincident artifact
see Section B.2.5.
2. Click the Apply button in the Analysis Parameters section.

To blank stimulation artifacts in files containing Electrical or Optical Stimulation Tags:

1. Click the Remove Spikes Post-Stimulation Tag checkbox.

2. Type the removal duration (ms) in the field. A removal duration of 2 ms is recommended, increasing if
needed, not to exceed 6 ms. Spikes detected during the removal duration after a tag will be removed.
3. Click the Apply button in the Analysis Parameters section.

To restore all analysis parameters to the default settings, select File  Restore Defaults.

B.2.2. Bursting Display and Parameters

The Bursting section displays a raster plot of the detected spikes on each electrode within the active well. Each
tick indicates the time a spike occurred and each row indicates the electrode. Blue ticks indicate the spikes are
part of a single-electrode burst while black ticks are not. Ticks included in network bursts are outlined by
magenta rectangles. Above the raster is a filtered population spike time histogram, the total number of spikes
occurring throughout the well at each time. When Electrical or Optical Stimulation Tags are present in the data
file (See Chapter 4), black triangles appear at the bottom of the plot to indicate tag timing.

To show/hide network burst detection:

1. Select/Deselect the Network Bursts checkbox.

Note: Network bursting will still be calculated, the checkbox only effects the data display.

To show/hide the filtered population spike time histogram:

1. Select/Deselect the Spike Histogram checkbox. Zoom in, zoom out, and zoom reset (+, -, =
respectively) are available for closer inspection of the histogram.

To change the time window displayed:

1. Type the display start and end time into the fields above the plot.
Note: This time window is for display purposes only. The analysis window is set in the Analysis
Parameters section.
2. Click the Refresh button to the right of the time fields.

[84]
 Neural Metric Tool

Show/Hide Network
Refresh Displayed Time
Bursts and Spike Histogram

Histogram
zoom controls Filtered Population
Copy Figure
Spike Time Histogram

Raster Plot

The Burst Parameters section defines single electrode and network bursting analysis parameters. Single
Electrode Bursts can be identified using the same algorithms available in AxIS, inter-spike interval (ISI)
threshold and Poisson surprise. Network Bursts can be identified using ISI threshold, Adaptive, or Envelope
algorithms. While AxIS identifies network bursting using the ISI threshold method, the Adaptive and Envelope
methods are only available in the Neural Metric Tool.

Many of the same metrics computed for burst detection on single electrodes are also computed for network
bursts (See Section B.4). The network burst metrics tend to be more robust, as they consider activity over the
entire network, but single electrode burst metrics will be more accurate when bursting behavior is independent
across electrodes within a well. In this way, single electrode burst metrics and network burst metrics are
complementary, and their use will depend on the cell type/source and characteristics of the network activity.

To set single electrode burst detection:

1. Select ISI Threshold or Poisson Surprise from the Single Electrode Bursts drop-down list:
1.1. ISI Threshold: Bursting is defined as at least N spikes on an electrode, each separated by an
inter-spike interval (ISI) of no more than T seconds. The method is adapted from Chiappalone et
al., 2005. Set the minimum number of spikes (N) and maximum time (T) between each spike
using the Min # of Spikes and Max ISI (ms) fields, respectively.
1.2. Poisson Surprise: Assumes the neurons are firing according to a Poisson distribution. It assesses a
collection of spikes and determines how improbable it is a chance occurrence according to a
“surprise” threshold. The method is adapted from Legéndy & Salcman, 1985. In this way, the
algorithm is adaptive to the mean firing rate on each electrode. The Min Surprise field

[85]
AxIS 2.4 User Guide

determines how sensitive burst detection is, such that a low Min Surprise will identify bursts more
frequently.
2. Click Apply in the Burst Parameters section.
Note: To exclude electrodes that don’t meet a minimum bursting rate see Section B.2.5.

To set network burst detection:

1. Select ISI Threshold, Adaptive, or Envelope from the Network Bursts drop-down list:
1.1. ISI Threshold: Defines a network burst as a collection of at least N spikes across all electrodes in the
well, each separated by an inter-spike interval of no more than T seconds with at least X percent of
electrodes participating in the burst. Adapted from Bakkum et al. 2013. Set the minimum number of
spikes (N), maximum time between each spike (T), and minimum electrodes (X) using the Min #
of Spikes, Max ISI (ms), and Min Electrodes (%) fields, respectively.
1.2. Adaptive: Defines a network burst the same as ISI Threshold. The maximum time between spikes
(Max ISI) is set automatically on a well-by-well basis based on the mean firing rate of each
well; wells with a higher mean firing rate have a lower Max ISI. In this way, the identification of
network bursts is not biased by tonic activity in the well. Set the minimum number of spikes and
minimum electrodes using the Min # of Spikes and Min Electrodes (%) fields, respectively.
1.3. Envelope: Detects network bursts based on the filtered population spike time histogram, which is
created by applying a Gaussian window to the binned spike times across the well. The algorithm
defines a network burst by identifying times when the histogram exceeds a threshold of N standard
deviations above or below the mean. Bursts must be separated by T seconds and include at least X
percent of electrodes. Set the number of standard deviations (±N), minimum interburst interval (T),
and percent of electrodes (X) with Threshold Factor, Min IBI (ms), and Min Electrodes (%),
respectively. The boundaries of the network bursts are defined when the histogram falls back to near
baseline. Burst Inclusion (%) defines how spikes near the boundaries are included with higher values
including more spikes.
2. Click Apply in the Burst Parameters section.
Note: To exclude wells from network bursting calculations see Section B.2.5.

B.2.3. Synchrony Display and Parameters

The Synchrony section displays either the standard or normalized well-wide cross-correlogram. A cross-
correlogram describes how closely related two spike trains are. The well-wide cross-correlogram uses
frequency domain methods (Halliday, Rosenber, Breeze & Conway 2006) to compute and pool the cross-
correlogram across all unique pair-wise combinations of electrodes in a well. The normalized cross-
correlogram normalizes the inter-electrode cross-correlations by their auto-correlations. This normalization
reduces the effects of single electrode high mean firing rate and/or high spiking regularity which can
artificially increase correlations. The vertical gray lines indicate the size of the synchrony window.

[86]
 Neural Metric Tool

To change the display between the standard and normalized cross-

correlograms:

1. Click either the Cross-Correlogram or Normalized Cross-Corr for

standard cross-correlogram or normalized cross-correlogram,
respectively.

The Synchrony Parameters section defines synchrony analysis parameters. The

Synchrony Window (ms) field is the window of time around zero that is used
to compute the area under the cross-correlation, and area under the
normalized cross-correlation. A short synchrony window (e.g. 5 ms) quantifies
synchrony on a millisecond timescale, while a long synchrony window (e.g. 100
ms) captures synchronous activity on slower timescales.

The Neural Metric Tool computes some synchrony metrics automatically (See Section B.4), but can also
compute the Kreuz metric (derived from the Kreuz spike distance, Kreuz et al., 2013) and the Synchrony
Index (based on Paiva, Park, and Principe, 2010).

To calculate the Kreuz Metric and Synchrony Index:

1. Select Kreuz Metric, Synchrony Index, or All from the Additional Synchrony Metrics dropdown.
2. Click the Apply button in the Synchrony Parameters section.

B.2.4. Evoked Display and Parameters

The Evoked section only displays data when Electrical or Optical Stimulation Tags are present in the recording
(see Chapter 4). The peri-stimulus histogram (top) displays the aggregate well-wide response averaged
across the stimulus repeats for the active well. The peri-stimulus raster (bottom) displays the response of each
electrode to the stimulus, binned by time, and averaged across the stimulus repeats. Tick darkness increases as
spike number increases for that bin.

Peri-stimulus Histogram

Peri-stimulus Raster

[87]
AxIS 2.4 User Guide

The Stimulation Parameters section defines evoked activity analysis parameters. Use the fields in the
Stimulation Parameters section to specify which stimulus Events to analyze (set as a range of stimulus
numbers), the time Window relative to these events to use for analysis, and the Bin Size used to aggregate
spikes for the peri-stimulus raster and peri-stimulus histogram.

B.2.5. Advanced Options

The Advanced menu option under File provides options for spike blanking after coincident artifacts, setting
burst activity criteria, and selecting wells for network burst exclusion.

To blank spikes after a coincident artifact:

1. Click File Advanced Artifact Blanking.

2. Type the blank duration (ms) in the Blank field in the Analysis Parameters section.
3. Click the Apply button in the Analysis Parameters section. Spikes detected during the blank duration after
each coincident artifact will be removed.

To apply a minimum bursting criteria:

1. Click File  Advanced  Burst Criterion.

2. Type a minimum bursting rate (bursts/min) in the Burst Electrode Criterion field in the Burst Parameters
section.
3. Click the Apply button in the Burst Parameters section.
Note: Electrodes with a burst rate below the minimum burst rate will be excluded from electrode burst
metrics for both Single Electrode Measurements and Well Averages.

To exclude wells from network burst calculations:

1. Click File Advanced  Turn off Network Bursts in select wells.

2. Click the well(s) in the Network Bursting window to disable detection of network bursts in those
well(s). White wells are enabled, grey wells are disabled.
3. Click Apply to recalculate network bursting using only the enabled wells.
Note: Click Reset to turn on all wells, click Restore Defaults restore all wells to the state they were in
when the Network Bursting window was opened.

[88]
 Neural Metric Tool

B.3. OPERATION

B.3.1. Generate AxIS Spike file

The Neural Metric Tool uses AxIS Spike (.spk) files generated by AxIS. While plate map information is not
used in the Neural Metric Tool, the information is preserved in the .csv output for use in other tools. Always
enter complete plate maps and experiment notes to an AxIS Raw file prior to beginning analysis (See Section
2.2.1). Follow these steps to create AxIS Spike files in AxIS.

1. For a single file:

1.1. Select File  Open Recording….
1.2. Select the file for analysis and click Open.
2. For multiple files:
2.1. Select File  New Batch Process….
2.2. Click Add in the Edit Batch Process Settings dialog.
Note: Use Remove to remove any files from the Source File list.
2.3. Select the desired .raw files and click Open.
2.4. Select the analysis window using the Segment Type dropdown menu. See Section 5.4 for setting
an analysis window with the Segment Type dialog.
2.5. Click OK.
3. Right-click on the file or batch process in the Streams window and select Configuration  Neural Offline
 Spontaneous, Electrically Evoked, or Optically Evoked.
Note: Select Spontaneous for most applications, Electrically Evoked if electrical stimulation was used
during the recording, or Optically Evoked if optical stimulation was used during the recording.
4. Right-click on the Burst Detector in the Streams window and select Remove.
Note: This step will increase analysis speed but disable generation of file outputs dependent on the Burst
Detector and Neural Statistics Compiler. If using file outputs from these processors, skip this step.
5. Click on the Experiment Setup Properties module.
6. Select AxIS Spike from the Spike Detector dropdown.
7. Optional: Uncheck Auto Name File beside the selected file outputs to manually enter a file name. By
default the name will be [SourceFile]_[StreamName].
8. Click Record or Start Batch Process. The AxIS Spike files are saved to the AxIS Raw file directory.

Select AxIS Spike

[89]
AxIS 2.4 User Guide

B.3.2. Analyze a single AxIS Spike file

To analyze a single AxIS Spike file in the Neural Metric Tool, follow these steps:

1. Click File Load Axion Spike File….

2. Select the file for analysis and click Open. Depending on the length of the recording, it can take several
minutes to load the data.
Note: The window header now displays the loaded file name. The Plate Map section displays the selected
metric across wells. The Bursting, Synchrony, and Evoked sections display a single well. The Evoked
section will only display data if electrical or optical stimulation tags are present in the file (see Chapter 4).
3. Adjust the analysis settings using the Analysis Parameters, Burst Parameters, Synchrony Parameters, and
Evoked Parameters sections. See Section B.2for analysis setting options.
Note: To restore analysis settings to their default values, click File  Restore Defaults.
4. Export desired figures:
4.1. Click the target well on the Plate Map to make it the active well.
4.2. Adjust any display options, see Section B.2 for display options.
4.3. Click the Copy button beside the desired figure.
4.4. Click Save As in the figure window.
4.5. Type a file name, select a file extension, and click Save.
Note: Clicking Copy also copies the figure to the clipboard to directly paste into other software programs.
5. Export calculated metrics:
5.1. Click File  Export  Export Recommended Metrics to CSV, Export Supplemental Metrics to CSV,
or Export Metrics to Matlab File. See Section B.4 for file output types.
5.2. Type a file name and click Save.

B.3.3. Analyze multiple AxIS Spike files

To analyze a multiple AxIS Spike files in the Neural Metric Tool in a batch process:

1. Adjust the analysis settings using the Analysis Parameters, Burst Parameters, Synchrony Parameters, and
Evoked Parameters sections. See Sections B.2 for analysis setting options.
Note: Analysis settings may be altered in the Neural Metric Tool window at any time. Click Update
Neural Metric Tool Parameters in the Batch Process window to analyze with updated settings.
2. Click File  Batch process multiple files.
3. Click Add Files in the Batch Process window.
4. Select the AxIS Spike files to analyze and click Open.
Note: To exclude files from analysis after opening, deselect it from the table in the Batch Process window
using the Select checkbox. Click Clear Files to remove all files from the batch process.
5. Select the analysis window from the Analysis Parameters drop-down menu.

[90]
 Neural Metric Tool

Segment Type Description

Whole File Analysis window is the entire duration of the file.
Start of File Analysis window begins at the start of the recording
plus the Offset and continues for the Duration.
End of File Analysis window begins at the end of the file minus
the Duration and continues until the end.
6. Select the file output (Export Recommended Metrics to CSV, Export Supplemental Metrics to CSV, or
Export Metrics to Matlab File) from the Export Options drop-down menu.
7. Optional: Type a file name suffix to append to output files in the Enter File Suffix field.
8. Click Run Batch.
Note: Exported neural metric files will be saved to the path of the source spike file and automatically
named [spike file name][_neuralMetrics][_suffix][.file extension].

Selected Files
Loaded Files

Output File Suffix

Analysis Window

Start Batch

B.4. OUTPUT
The Neural Metric Tool can export calculated metrics to a .csv file or a .mat file. The resulting .csv file can
be opened with any software that can read a text file, such as Microsoft Excel. The Neural Metric Tool
output mirrors the output of the Advanced Metrics output generated by the Neural Statistics Compiler in AxIS
in layout (See Section 7.4). Unlike the Advanced Metrics output, treatment group averages are not included in
the Neural Metric Tool output.

The output metrics of Export Recommended Metrics to CSV (recommended) and Export Supplemental Metrics
to CSV (supplemental) are listed in the table below. In general recommended metrics contain the mean values
while supplemental metrics contain mean and median values. A few additional non-median metrics are
included in supplemental metrics. Export Metrics to Matlab File exports the supplemental metrics to a .mat file
for further custom analysis.

[91]
AxIS 2.4 User Guide

Recommended Metrics Supplemental Metrics

Activity Metrics Activity Metrics
Number of Spikes* Number of Spikes*
Mean Firing Rate* Mean Firing Rate*
ISI Coefficient of Variation* ISI Coefficient of Variation*
Number of Active Electrodes* Network ISI Coefficient of Variation
Weighted Mean Firing Rate* Number of Active Electrodes*
Electrode Burst Metrics Weighted Mean Firing Rate*
Number of Bursts* Electrode Burst Metrics
Number of Bursting Electrodes* Number of Bursts*
Burst Duration* Number of Bursting Electrodes*
Number of Spikes per Burst* Burst Duration (Mean and Median)*
Mean ISI within Burst* Number of Spikes per Burst (Mean and Median)*
Median ISI within Burst* Mean ISI within Burst*
Median/Mean ISI within Burst Median ISI within Burst*
Inter-Burst Interval* Median/Mean ISI within Burst
Burst Frequency* Inter-Burst Interval (Mean and Median)*
IBI Coefficient of Variation* Burst Frequency*
Burst Percentage* Normalized Duration IQR*
Network Burst Metrics IBI Coefficient of Variation*
Number of Network Bursts* Burst Percentage*
Network Burst Frequency* Network Burst Metrics
Network Burst Duration* Network Bursts Ignored Flag
Number of Spikes per Network Burst* Number of Network Bursts*
Mean ISI within Network Burst Network Burst Frequency*
Median ISI within Network Burst Network Burst Duration (Mean and Median)*
Median/Mean ISI within Network Burst Number of Spikes per Network Burst (Mean and Median)*
Number of Elecs Participating in Burst* Mean ISI within Network Burst
Number of Spikes per Network Burst per Channel* Median ISI within Network Burst
Network Burst Percentage* Median/Mean ISI within Network Burst
Network IBI Coefficient of Variation* ISI CoV within Network Burst
Network Normalized Duration IQR* Number of Elecs Participating in Burst (Mean and Median)*
Number of Spikes per Network Burst per Channel (Mean and
Synchrony Metrics
Median)*
Area Under Normalized Cross-Correlation* Network Burst Percentage*
Area Under Cross-Correlation* Network IBI Coefficient of Variation*
Full Width at Half Height of Normalized Cross-Correlation Network Normalized Duration IQR*
Full Width at Half Height of Cross-Correlation* Synchrony Metrics
Synchrony Index* Area Under Normalized Cross-Correlation*
Kreuz SPIKE Distance Area Under Cross-Correlation*
Evoked Metrics Full Width at Half Height of Normalized Cross-Correlation
Number of Trials Full Width at Half Height of Cross-Correlation*
Evoked Spike Count Synchrony Index*
Evoked Response Probability Kreuz SPIKE Distance
Evoked First Spike Latency Evoked Metrics
Evoked Jitter Number of Trials
Evoked Spike Count
Evoked Response Probability
Evoked First Spike Latency
Evoked Jitter
Metrics marked with an asterisk (*) in the table above are defined in Section 7.4. Additional metrics provided
by the Neural Metric Tool are defined below.

[92]
 Neural Metric Tool

Metric Definition
Average across electrode bursts of the median/mean inter-spike interval (ISI) within
Median/Mean ISI within Burst electrode bursts. Values close to 1 indicate the distribution of ISIs within bursts is
symmetric.
Mean ISI within Network Burst Average across network bursts of the mean ISIs within network bursts.
Median ISI within Network Burst Average across network bursts of the median ISIs within network bursts.
Average across network bursts of the median/mean ISI within network bursts. Values
Median/Mean ISI within Network Burst
close to 1 indicate the distribution of ISIs within bursts is symmetric.
Distance along the x-axis (phase lag) from left half height to right half height
Full Width at Half Height of Normalized (probability) of the normalized cross-correlogram. This is a measure of network
Cross-Correlation synchrony; higher half widths indicate a wider correlogram (less synchrony) whereas
lower half widths indicate a taller correlogram (greater synchrony).
1-Kreuz SPIKE distance (Kreuz et al 2013) such that 1 is perfect synchrony and 0 is
perfect asynchrony. It is computed with only one previous and one subsequent spike
Kreuz SPIKE Distance for each reference spike. The time window for each computation varies with local
firing rate of each spike train. It tracks changes in instantaneous clustering without
being skewed by individual electrode inter-spike interval.
Number of Trials Number of stimulation events used for analysis.
Average across trials of the number of spikes detected across all electrodes in the well
Evoked Spike Count during the time window specified by the Stimulation Parameters section. This is a
measure of the magnitude of stimulus response.
Probability of finding at least one spike in the well during the time window specified
by the Stimulation Parameters section. Computed as the number of stimulation events
Evoked Response Probability
that evoked a response divided by the total number of stimulation events. A measure
of the response to the stimulus.
The average across trials of the time between the stimulation event and the first post-
Evoked First Spike Latency
stimulus spike detected in the well.
The standard deviation across trials of the time between the stimulation event and the
Evoked Jitter first post-stimulus spike detected in the well. This is a measure of response consistency;
lower values indicate more consistent responses.
Coefficient of variation (standard deviation/mean) of the inter-spike interval for all
spikes on all electrodes in a well. A measure of spike regularity. This metric captures
the distribution of spiking such that 0 indicates spikes perfectly distributed and > 1
Network ISI Coefficient of Variation indicates network bursting; as network bursting becomes clearly distinguished from
quiescence across the network, the standard deviation of the ISI grows because
spikes inside a burst have low ISI but there are large ISIs between the last spike in one
burst and the first spike in the next burst.
A flag to indicate whether network bursts were ignored in this well. Network bursts
Network Bursts Ignored Flag
ignored = 1, network bursts included = 0.
Average across network bursts of the ISI CoV (standard deviation/mean of the inter-
ISI CoV within Network Burst
spike interval) within network bursts.

[93]
AxIS 2.4 User Guide

APPENDIX C. CIPA ANALYSIS TOOL

C.1. INTRODUCTION
The CiPA Analysis Tool is a comprehensive cardiac analysis software designed to be suitable for any cardiac
analysis application requiring precise assessment of field potential duration (FPD) and arrhythmia. The CiPA
Analysis Tool requires AxIS 2.3 or greater.

The CiPA Analysis Tool operates according to the workflow described in the figure below. AxIS identifies all
beats and a stable period of beating for each well in Stage 1. In Stage 2, a “Golden Channel” is identified,
which is an electrode that has a repolarization feature that is trackable across all conditions. Finally, FPD is
calculated and arrhythmic events are identified and classified. The user reviews the results of the automated
algorithm and has the option to manually correct the repolarization timing measurements.

Stage 1 Stage 2 Stage 3

Beat Period “Golden FPD Calculation

Stability Analysis Channel” and Arrhythmia
(using AxIS 2.3) Identification Classification

C.2. CIPA ANALYSIS TOOL OVERVIEW

The CiPA Analysis Tool has three main sections: the Golden Channel Selector, FPD Detection Display, and the
Arrhythmia Inspector. The Golden Channel Selector is on the left side of the main screen while the FPD
Detection Display is on the right. The Arrhythmia Inspector is accessible through the FPD Detection Display.

Golden Channel Selector FPD Detection Display

Access Arrhythmia Inspector

[94]
 CiPA Analysis Tool

C.2.1. Golden Channel Selector

The Golden Channel Selector is used to select the electrode to use for FPD analysis. The waveforms from the
selected electrode are displayed in FPD Detection window. A “Golden Channel” is an electrode with a clear,
consistent beat waveform. The T-wave should be identifiable across files and its timing representative of the
electrodes in the well.

The beat waveforms from each file are displayed on a single plot for each electrode in the left panel. The
displayed waveform is an average of 5 beats from the stable region identified by AxIS. The tool selects a
“Golden Channel” by identifying the electrode with the largest T-wave in the baseline file and highlights the
electrode in yellow.

To select a different “Golden Channel” electrode:

1. Click on the plot of another electrode.

1.1. Optional: Adjust the x- and y-axis to view the beat waveform using the x axis max and y axis max
fields, respectively.

Selected Electrode

Current Well
Export Figures

Plot Limits Exclude Well

Click Export to generate two additional plots for the selected electrode. The first is a larger, interactive copy of
the plot shown in the tool display. The second is an interactive plot showing the individual beats used to
compute the displayed average beat for all conditions. These plots may aid in selection of the “Golden
Channel”.

Use the Exclude Well checkbox to exclude a well from analysis. The endpoints from this well will be excluded
from output figures and .csv reports (Section C.4).

Select a “Golden Channel” according to the following criteria:

1. The T-wave is present on each file.

2. The T-wave time is comparable to other electrodes.

[95]
AxIS 2.4 User Guide

3. A T-wave feature is similar between files.

Note: The FPD, amplitude, or beat period may increase or decrease between files but features like the T-
wave polarity should remain constant.

Example A: Electrode 33 has a larger T-wave in the

baseline file (black) than electrode 13, and thus would A
be chosen by default. The comparison file (red) T-
wave is not detectable in electrode 33, while it is for
electrode 13. For this reason, electrode 13 should be
chosen as the “Golden Channel”.
B
Example B: An electrode with a negative T-wave
provides the most reliable feature tracking between
files. Electrode 31 shows a strong positive T-wave in
the baseline file, but the comparison file T-wave is not
detectable. Conversely, electrode 23 exhibits a negative T-wave for repolarization, but is consistent and
easily tracked across the baseline and comparison files. In this case, electrode 23 should be chosen as the
“Golden Channel”.

C.2.2. FPD Detection Display

The FPD Detection Display is used to verify FPD measurements. It Selected FPD
shows the beat waveform across all files for the electrode selected
by the Golden Channel Selector. The tool will automatically
compute the peak of the T-wave for each file and display it as a
vertical line on the beat waveform. The FPD value is displayed in the
summary table beneath the plot.

If the incorrect T-wave is identified for any file, click on the correct
location on the beat waveform. The tool will auto-fit a new peak in a Plot Display
local region around the click.

Press Confirm to accept the FPD measurements and arrhythmia

classifications (if applicable, see Section C.2.3) and proceed to the
next well. Prior to confirming a well, if a new “Golden Channel” is Confirm
selected, the tool will recalculate the FPD based on the new selected
channel. After a well has been confirmed, the selected FPD values are retained for the well, regardless of
which electrode is selected.

The drop-down menu below the plot specifies how data is displayed in the plot. All of the displays allow click-
correct of the FPD. The plot display options are:

1. Stacked (default): The beat waveforms from each file are spaced vertically in the display. The T-wave
peak in each file is marked by a vertical line.

[96]
 CiPA Analysis Tool

2. Overlay: The beat waveforms from each file are plotted on top of one another. The T-wave peak in
each file is marked by a vertical line.
3. Display name (example: Baseline, Dosed, etc.): The beat waveforms for a single file from all
electrodes spaced vertically. The selected electrode in the Golden Channel Selector is displayed in
blue at the top. Behind the blue trace, the waveforms from the other files for the selected electrode are
displayed in gray. The “Golden Channel” T-wave peak is marked by a vertical line that crosses all
electrodes.

Stacked Overlay Display Name

C.2.3. Arrhythmia Inspector

The Arrhythmia Inspector displays beat period and continuous voltage plots. It highlights and displays regions
of instability and allows for the verification and classification of arrhythmic events.

To the right of the FPD Detection Display plot are buttons labeled with the Display Name for each file. If
arrhythmic events are detected in a file, based on a numerical algorithm looking for instability in beat period
from beat to beat, the button for that file will change to red. In these cases, the tool will automatically select
arrhythmia type A for comparison files.

Note: It is possible to have unstable beating when no arrhythmia is present, for example due to a drift in beat
period. Likewise it is possible to have stable beating when arrhythmic events are present, particularly if every
beat contains the EAD feature. Thus, the red button should only be used as a guide and all arrhythmic events
should be visually verified and classified.

Clicking one of the buttons will open the Arrhythmia Inspector display for that file.

The top left plot displays beat period over time, showing all beats included in the .csv file from AxIS. The top
right plot displays a Poincare plot of the last five minutes only. If unstable beating is detected, beats in that
region are displayed as red circles. The beats used to generate the average beat waveform for the FPD
Detection Display are indicated by green circles. The bottom plots display representative continuous voltage
from the electrode selected by the Golden Channel Selector. The right Stable Beats plot displays data from the
stable beating region identified in AxIS. The left Unstable Beats plot displays data from a region of unstable
beating if one was identified. Use the continuous voltage data plots to check for arrhythmic events or
quiescence.

[97]
AxIS 2.4 User Guide

If depolarization spike amplitudes were lower than the detection threshold in AxIS, the inspection button on
the main figure will change to bold and the well will be marked Quiescent, not beating. View the continuous
voltage plots in the Arrhythmia Inspector and if small amplitude beats are present in the raw data but were too
small to be detected by AxIS, leave the Quiescent box unchecked. If no beats are present, check the box
beside Quiescent to label this well/condition as quiescent in the output .csv file.

Sub-threshold beating (Not Quiescent) Quiescent

Note: Raw data is down-sampled by 10x in the CiPA Analysis Tool to speed computation and save time.
Amplitudes displayed in the continuous voltage plots may not be accurate. The down-sampling only affects the
visualization of these plots. Amplitudes reported in the output .csv files and figures are calculated by AxIS,
which uses the complete dataset.

The bottom right Review dialog is used to mark and classify arrhythmic events. Each of the checkboxes
corresponds to a particular arrhythmia classification, as defined by the CiPA Myocyte Committee. Click
Legend to view an example of each arrhythmia classification.

Compare the examples of arrhythmic events to the continuous voltage data plotted in the Unstable Beats plot.
If more than one type of arrhythmia is present, select all of the corresponding checkboxes. Classify any
arrhythmic activity that occurs, even if it does not occur during the time window used for analysis. Use the
Notes field to enter any interesting observations about the well. Press Confirm to accept selections and return
to the main window.

[98]
 CiPA Analysis Tool

C.3. OPERATION

C.3.1. Process Recordings in AxIS

Continuous recording is required to take advantage of the full utility of the analysis tools. The CiPA Analysis
Tool utilizes the Advanced Metrics .csv output from the Cardiac Statistics Compiler in AxIS. Follow these steps
to process the data using AxIS 2.3 or above:

1. Enter the plate map information into the baseline .raw file if it is
not already present. See Section 2.2 for additional information
about plate maps.
2. Click File  New Batch Process...
3. Click Add in the Edit Batch Process Settings dialog.
4. Select the desired .raw files and click Open.
5. Select Whole File in the Segment Type drop-down menu.
6. Click OK.
7. Right-click on the batch process in the Streams window and
select Configuration  Cardiac Offline  Spontaneous.
8. Double-click the Cardiac Beat Detector to open the settings and modify
them if desired. Recommended settings:
Detection Threshold = 300 µV
Min Beat Period = 250 ms
Max Beat Period = 10 s
FPD Method: Polynomial Regression
Note: It is not necessary to optimize FPD detection settings in AxIS
because the CiPA Analysis Tool allows adjustment of FPD if
required. However, the FPD reported by AxIS is used as an initial
condition in the CiPA Analysis Tool.
9. Double-click the Cardiac Statistics Compiler to open the settings.
Recommended Settings:
Limit to Region of Most Stable Beat Period = Enabled
Included Source Data = All Beats
Set the analysis window using the Region Limit, Duration, and
Offset fields.
10. Confirm Advanced Metrics is selected from the Statistics Compiler
drop-down menu in Experiment Setup Properties.
11. Click Start Batch Process to run the batch process and save the .csv
files. The files are saved to the same directory as the .raw files.

[99]
AxIS 2.4 User Guide

C.3.2. Analyze Data in the CiPA Analysis Tool

Loading a new experiment will take several minutes. The software computes an average beat waveform for
every electrode in every file loaded. Status bars will display loading progress. After the data has been loaded
and processed, new .mat files are created corresponding to each .raw file selected for analysis. Do not delete
these .mat files, as they allow much quicker loading in the future.

1. Ensure the .raw files and associated .csv files are in the same directory.
2. Click File  Load  New Experiment. This opens a new File Selector window.

Select Baseline Files Select Comparison Files

3. Click Baseline to import a baseline file that will be used as the reference point for the comparison files.
4. Navigate to the directory where the files are saved, hold Ctrl and select both the baseline .raw and .csv
files.
5. Click Comparison to import up to four files to compare to the baseline file.
6. Navigate to the directory where the files are saved, hold Ctrl and select a .raw file and then a .csv file for
each comparison file.
7. Optional: Adjust the file order. Comparison files will be loaded in alphabetical order.
7.1. Type numbers into the File Order column. Files will be displayed in order from least to greatest.
8. Optional: Adjust the display name.
8.1. The default display name for the baseline file is “B”, and the comparison files are “C1”,”C2”,”C3”,
etc.
8.2. Type a new name into the Display Name column. Names longer than 5-8 characters may not
display properly on exported figures.
9. Click “Confirm” to apply the changes.

[100]
 CiPA Analysis Tool

10. Optional: Load a new plate map if there is no plate map in the file or to use an alternate plate map. If
there is a plate map in the baseline .csv file it will automatically be used for analysis.
10.1. Click File  Load  New Plate Map.
10.2. Navigate to and select a .platemap file or .csv file and click Open. For more information about
generating plate maps in AxIS and saving .platemap files, see section 2.2.
11. Select a “Golden Channel” in the Golden Channel Selector. See Section C.2.1.
11.1. Click on the desired electrode plot if not already highlighted in yellow.
12. Verify the T-wave peak detection in the FPD Detection Display. See Section C.2.2.
12.1. Click the T-wave peak on the beat waveform to correct any misidentified T-waves.
13. Inspect for arrhythmia in the Arrhythmia Inspector. See Section C.2.3.
13.1. Click on the display name button to the right of the FPD Detection Display.
13.2. Click on any relevant checkboxes in the Review dialog.
13.3. Click Confirm.
14. Click Confirm to proceed to the next well.
15. Repeat steps 11-14 for each well.
Note: If a well was skipped or confirmed prematurely, select it in the Well drop-down menu above the
Golden Channel Selector.
16. Click File  Save Endpoints to .mat
17. Type a file name for the .mat file and click Save.
Note: An incomplete analysis can be saved and exited at any time for future completion. The saved .mat
file contains the file names, display names, Golden Channel selections, FPD selections, and arrhythmia
classifications.
18. Click the Export button near any figure to copy the image to the clipboard and save the figure as a variety
of file types.
19. Click File  Export  Export Figures, Export Plot Source Data to CSV, Export Well Endpoints to CSV, or
Export CiPA CSV to create file outputs. See Section C.4 for more information about the CiPA Analysis Tool
outputs.

C.3.3. To load a saved experiment

Previously loaded experiments may be reloaded for continued analysis.

1. Click File  Load  Saved Experiment.

2. Navigate to and select the .mat file saved previously. If the name of any .raw or .csv files associated with
the saved experiment have changed, the .mat file will not load properly.

[101]
AxIS 2.4 User Guide

C.4. OUTPUTS

C.4.1. Export Figures

Export Figures creates bar plots for FPD, FPDc (FPD corrected for Beat Rate using Fridericia’s correction), Beat
Period, and Spike Amplitude. The endpoints for each well are first computed as a percentage change from the
baseline file, and replicate wells are averaged to generate the dose-response. The error bars show the
standard deviation across replicate wells. The number of wells containing arrhythmic events for each condition
are indicated on the plots by an asterisk and a fraction indicating the number of arrhythmic wells out of the
total number of wells in that condition. If there is only one comparison file, each treatment from the plate map
will be plotted on a separate plot with concentration on the x-axis. If there are multiple comparison files, each
group (treatment + concentration) from the plate map will be plotted on a separate plot with comparison files
on the x-axis.

The figures also include summary plate map plots for each metric and Beat Period CoV. On these plots, *
indicates arrhythmia in that well, ○ indicates no beats were detected in AxIS, and ● indicates beating was
detected, but no FPD was identified (usually due to a flat T-wave or arrhythmic event).

[102]
 CiPA Analysis Tool

C.4.2. Export Plot Source Data to CSV

Export Plot Source Data to CSV exports the data used to generate the replicate-averaged dose-response plots
shown in the exported figures as a .csv file. For each endpoint, the average and standard deviation across
replicates of the percent change from baseline are provided for each comparison file, as well as the individual
replicate values that went into the average.

C.4.3. Export Well Endpoints to CSV

Export Well Endpoints to CSV exports the cardiac metrics for each well to a .csv file. For each file, a block of
statistics are provided, organized according to well. The endpoints from all analyzed files are saved to a
single .csv file.

C.4.4. Export CiPA CSV

Export CiPA CSV exports the cardiac metrics for each well to a .csv
file formatted to the reporting requirements as defined by the CiPA
Myocyte Committee. Enter experiment information in the dialog box
that appears. Type the test site code, cell type code, platform code,
and plate number into the respective fields, and click Confirm.
Endpoints are saved in a row-based format. The endpoints from all
analyzed files are saved to a single .csv file.

[103]
AxIS 2.4 User Guide

APPENDIX D. AXIS METRIC PLOTTING TOOL

D.1. INTRODUCTION
The AxIS Metric Plotting Tool allows rapid visualization of experiment results and organization of endpoints
according to treatment condition. This tool imports all endpoints generated by AxIS, along with a plate map,
and calculates % change relative to baseline for each well, averages replicates, and generates plots
comparing treatment conditions or time points. The tool allows for viewing of different endpoints with a single
click, facilitating data exploration.

D.2. OPERATION

D.2.1. Load files for analysis

1. Import an Advanced Metrics or Neural Metrics .csv file that will be used as the reference point, or
baseline, for the comparison files:
1.1. Click Baseline….
1.2. Select the baseline .csv file and click Open.
2. Import files to compare to the baseline file:
2.1. Click Comparison….
2.2. Select the comparison .csv file(s) and click Open.
Note: Hold Ctrl to select multiple files. Comparison files will
be loaded in alphabetical order.
3. Load a plate map.
3.1. If the baseline file contains a plate map it will be loaded automatically. Skip to Step 4.
3.2. Click Plate Map….
3.3. Select an Advanced Metrics .csv file or a .platemap file exported from AxIS.
4. Optional: Type numbers in the Plot Order column of the File Load table to change the order of the
comparison files. Baseline is 0, comparison files start at 1. To see the complete file names, hold the cursor
over the table and the full file names will appear in the tooltip.
5. Optional: Type names for each file in the Display Name column of the File Load table. The name will
appear on plots in place of the file name.
6. Click Confirm to apply the changes.
Note: File display name and plot order may be changed at any time. Click Confirm to update plots after
any changes.

[104]
 AxIS Metric Plotting Tool

Load Files

Load Plate Map

Data Display

Organize Treatments

Click Clear to remove all files and data from the AxIS Metric Plotting Tool.

D.2.2. Adjust data display settings

The lower left section of the AxIS Metric Plotting Tool contains options to specify which data is displayed.

1. Select either One Treatment Per Plot or One File Per Plot in the Treatment/Display Options dropdown.
When One Treatment Per Plot is selected, each plot contains a different treatment, with each file on the x-
axis. When One File Per Plot is selected, each plot contains a different file, with each treatment on the x-
axis.
One Treatment Per Plot One File Per Plot

[105]
AxIS 2.4 User Guide

2. Click the checkboxes under the Display column in the Treatment/Display Options table to select or
deselect treatments to plot.
Note: Click Select All to display all the treatments or One
compound at a time to display all the concentrations of a single
compound at a time. Scroll through compounds with the arrows to
the right.
3. Optional: Type numbers in the Plot Order column of the
Treatment/Display Options table to change the order treatments
are displayed starting at 1.
4. Select SEM or STD from the Error Bars dropdown to display
standard error of the mean or standard deviation error bars,
respectively.
5. Select % change or raw values from the Plot Display dropdown to
display data normalized to baseline values
(baseline/comparison x 100) or raw baseline and comparison values.
Note: If % change is selected, each well is normalized to its own baseline value, and the average %
change across replicate wells is displayed.
6. Optional: Exclude wells that do not meet a minimum activity criterion. Use the well exclusion dropdown to
select the exclusion parameter and type the minimum value into the Exclude wells below field. Exclusion
parameters include:
6.1. For neural: Mean Firing Rate (Hz), Number of Active Electrodes, Burst Frequency – Avg (Hz), Burst
Percentage - Avg, Number of Network Bursts, Network Burst Frequency (Hz), or Network Burst
Percentage.
6.2. For cardiac: Number of Beats or Total Active Electrodes.
7. Click Update to apply the changes.
Note: Data display settings may be changed at any time. Click Updates to update plots after any
changes.

D.2.3. Inspect plots

1. Select an endpoint to view from the Metric dropdown.

Note: When the dropdown is selected, endpoints may be scrolled through using the arrow keys.
2. Optional: Adjust the y-axis using the Y min and Y max fields. Click Apply to rescale and Clear Limits to
auto-scale.

[106]
 AxIS Metric Plotting Tool

Select Metric Export Figure

Axes Options

D.2.4. Export Figures and Files

1. To export figures:
1.1. Click Copy to export the currently displayed figures.
1.2. Click File  Save As… in the figure window.
1.3. Select a file type from the Save as type drop-down menu.
1.4. Type a file name and click Save.
Note: Upon clicking Copy, the image is also copied to the clipboard and can be pasted directly into a
document or presentation.
2. To export a .csv file do one of the following:
2.1. Click File  Export  Export Supplemental Metrics to CSV to export all endpoints in a multi-table
format. Type a file name and click Save.
2.2. Click File  Export  Export Recommended Metrics to CSV to export selected endpoints in a multi-
table format. Type a file name and click Save.
2.3. Click File  Export  Export Statistics Format to export all endpoints as a machine-readable table.
Type a file name and click Save.

D.3. OUTPUT
The output .csv files can be opened with any software that can read text files, such as Microsoft Excel.
The output lists the files and settings used for analysis, followed by group means, standard deviations or
standard error of the means (as selected in the software), and replicates. Results are given as %change
or raw values as selected in the tool. The data is grouped by metric according to the plot layouts with each
column representing a group on the plot. If One File Per Plot is selected, each column is a different treatment
condition and each file gets a separate table. If One Treatment Per Plot is selected, each column is a different
file and each treatment gets a separate table. The example below shows a .csv organized by One Treatment
Per Plot.

[107]
AxIS 2.4 User Guide

Export Supplemental Metrics to CSV exports all metrics from the starting .csv files(s) to the new output format.
To select only the most commonly metrics use Export Recommended Metrics to CSV. The metrics exported are
listed in the table below:

Recommended Metrics
Neural Cardiac
Mean Firing Rate (Hz) Beat Period Mean (s)
Number of Active Electrodes FPD mean (ms)
Number of Bursting Electrodes Spike Slope Mean (V/s)
Burst Frequency – Avg (Hz) Spike Amplitude Mean (mV)
Burst Duration – Avg (s) Conduction Velocity Mean (mm/ms)
Normalized Duration IQR – Avg
IBI Coefficient of Variation – Avg
Burst Percentage – Avg
Network Burst Frequency (Hz)
Network Burst Duration – Avg (sec)
Network Burst Percentage
Network IBI Coefficient of Variation
Network Normalized Duration IQR
Area Under Normalized Cross-Correlation

Export Statistics Format exports the data into a machine-readable format for use with statistical software
programs. It generates a table with wells in rows and well information and endpoints in columns. All files are
included in a single table.

[108]
 Axion Data Export Tool

APPENDIX E. AXION DATA EXPORT TOOL

E.1. INTRODUCTION
The .raw file format, the continuous voltage data, can only be used within AxIS. The Axion Data Export
Tool is used to convert .raw files into comma-separated value (.csv) files for use with 3rd party software.
For 3rd party software that use other AxIS file output types; such as AxIS Spike, Spike List, or Beat List, it is
not necessary to use the Axion Data Export Tool.

It is important to note that the .csv file generated by Axion Data Export Tool is significantly larger than the
.raw file. Limit the data selected for export to only a few electrodes at a time and only export short
periods of time (30 to 60 seconds maximum).

E.2. OPERATION
Install the Axion Data Export using the instructions in Appendix A. Start the Data Export Tool by double-
clicking the icon on the desktop or in the Start menu.

1. Click File → Load .RAW File.

2. Select a .raw file and click Open.
3. Note: The window header now displays the loaded
filename and the Wells and Electrodes selection boxes
display the available wells and electrodes.
3. Select either Choose Wells and Electrodes or Enter a
List.
3.1. If Choose Wells and Electrodes is selected: Select
the desired wells and electrodes from the Wells
and Electrodes lists.
3.2. If Enter a List is selected: Type a list of the desired
electrodes in the field below. Use the format
[well.electrode] separated by commas. Example:
A1.55, A2.83, B2.11.
4. Type the start and end time (in seconds) into the Start Time and Stop Time fields, respectively.
5. Optional: Click Renumber output times to start at 0 to offset the exported timestamps to begin at zero
seconds. If this box is not checked, the exported times will match the times in the original data.
6. Click Export....
7. Enter a file name and click Save to begin file export.

E.3. OUTPUT
The resulting .csv file can be opened with any software that can read a text file, such as Microsoft Excel.
The first column contains the timestamps of the data in that row. Each of the remaining columns contains
the voltage measurements for a single electrode at the time indicated in the first column. The exported file

[109]
AxIS 2.4 User Guide

is significantly larger than the original .raw file and much larger than most common spreadsheet files.
Some programs limit the number of rows and columns that a file can contain. Keep these limits in mind
when choosing the number of electrodes and the time points to export. It is usually best to choose the
desired times, wells, and electrodes by viewing the data in AxIS first.

[110]