Scribd
Upload
4 views

Kshama Presentation Final

The document outlines a project focused on developing an AI-based system for diagnosing Diabetic Retinopathy (DR) using deep learning techniques. It compares existing diagnostic methods with the proposed system, which is more accurate and time-efficient, and details the system's architecture, requirements, and testing results. The project aims to enhance accessibility and accuracy in DR detection, with future enhancements including a larger system and potential mobile app development.

Uploaded by

flashyellow81
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
4 views

Kshama Presentation Final

The document outlines a project focused on developing an AI-based system for diagnosing Diabetic Retinopathy (DR) using deep learning techniques. It compares existing diagnostic methods with the proposed system, which is more accurate and time-efficient, and details the system's architecture, requirements, and testing results. The project aims to enhance accessibility and accuracy in DR detection, with future enhancements including a larger system and potential mobile app development.

Uploaded by

flashyellow81
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 19

DIABETIC RETINOPATHY

IMAGE RECOGNITION USING


AI TECHNIQUE

Name : Kshama Devi K


Reg. No. : 832100426
Class : II MSc CS
Guide : Mr. R. Karthikeyan
Head of Computer Science
Department.
1. ABSTRACT
 Diabetic Retinopathy (DR) is caused by diabetes mellitus,
which can result in loss of vision.
 Thus diagnosing the disease before the vison loss can help
in retaining the vision.
 Here we aim in development of a system that uses
techniques of deep learning to identify whether the person is
affected or not by diabetics.
2. INTRODUCTION
The system will help in monitoring and detecting the diabetic
retinopathy thus helping in retaining the vision more accurately.
Module Description
1. Manual : It contains combination of different layers.

2. AlexNET : It Contains 5 convolutional layers, 3 max-pooling layers, 2


normalization layers, 2 fully connected layers, and 1 softmax layer.

3. LeNET : It consists of 2 convolutional layers, 2 max-pooling layers, 2


fully connected layers.

4. Deploy : It deals with the front-end implementation such as web page


designing using html and css languages.
About the company

Spiro Prime Tech Services is a professional IT service


organization offering world class IT development
services to clients. It is located at T. Nagar Chennai
where many professional expertise work together to
develop more technical society.
3. SYSTEM STUDY
Existing system
 Today the commonly used system is by detailed inspection
by the specialists.
 But they are expensive, time consuming and less accurate.
Proposed system
Here we propose a system which uses techniques of deep
learning to classify the image as affected or unaffected eye
that is more accurate and less time consuming.
Feasibility study
The system is operationally, technically and economically
feasible
4. BUSINESS DIAGRAM
5.SYSTEM ANALYSIS
5.1 DFD DIAGRAM

LEVEL 0

LEVEL 1
LEVEL 2
6. SYSTEM REQUIREMENTS
Hardware Requirement Software Requirement

 Processor: Intel Core™ I5- 8400 • Operating System: Windows 10

 CPU: 2.80GHzs • Front End: html,CS

 Memory: 512 SSD • Programming language: Python

 Mouse: DELL • Software used: Anaconda navigator,


Jupyter notebook, Pycharm,VS
 RAM:8 GB Code

 KeyBoard:110 keys enhanced


7. SYSTEM DESIGN
7.1 DATA COLLECTION

• Data is collected from www.kaggle.com .


• Size of the data set:
• Training: 1000 images
• Testing: 200 images
• Categories of images: 1. Non diabetic retinopathic
2. Severe diabetic retinopathic
• Format and size of images: Any image format such as Jpeg,
jpg of 30kb-70mb which is taken by Optical coherence
tomography (OCT). With this test, pictures provide cross-
sectional images of the retina that shows the thickness of the
retina. These images are fed in to the system to detect
diabetic retinopathy.
SYSTEM DESIGN
7.2 FORM DESIGN
CODING
EXAMPLE :
import os
import glob
import numpy as np
import pandas as pd
from PIL import Image
import numpy as np
import matplotlib.pyplot as plt
from tensorflow.keras.preprocessing.image import
ImageDataGenerator
8. TESTING
For testing, let us consider the case as user tries to upload
the image without selecting the file, we expect the system
to show text reminder to choose the file.
9.SCREENSHOT (WITH
VALUES )
10. Report

Manual architecture Accuracy-0.66 -

AlexaNet Architecture Accuracy – 0.54


LeNet Architecture _Accuracy – 0.95

Highest accuracy is obtained in LeNet Architecture.


Hence it is used in the project
11. CONCLUSION
System helps people in testing for diabetic
retinopathy by saving time and money. It
gives a bird eye view about how deep
learning architectures can be
implemented to predict diabetic
retinopathy.
12. FUTURE ENHANCEMENT
1. System be developed in a bigger version so that
eye test appointment taking, report generation,
patient details tracking and all such works done in
the same website.

2. This can be developed as an android app too, so


that it extends the accessibility.

3. More research scope is there in this project so that


more manual architectures can be build and
implemented for higher accuracy.

4. We can deploy this model in any cloud based


system.
Thank you

You might also like