Classification of Images Diagnosis of Brain

Diseases

1
 Contents
• Introduction
• Literature Review
• Problem Formulation
• Proposed Methodology
• Results and Discussions
• Conclusions and Future Scope

2
 Introduction
• Brain diseases significantly affect quality of life, emphasizing the importance of early detection and
treatment.
• Traditional imaging techniques like X-rays, MRI, CT, PET, and SPECT assist in diagnosing brain diseases.
• Accurate interpretation of medical images is challenging; radiologists may miss subtle disease indicators.
• Machine learning offers promise for brain disease identification and classification.
• Challenges include:
 Dataset diversity: Gathering varied and representative datasets is crucial.
 Feature extraction: Capturing relevant, subtle features from images is complex.
 Interpretability: Ensuring models' explanations for predictions are comprehensible.
 Generalizability: Models should perform well across different healthcare settings.
 Ongoing research is necessary to address these obstacles and improve machine learning applications in
brain disease diagnosis.
3
 Literature Review

S.No. Name of the Research Paper Name of the authors Summary

1. Classification and Visualization of Qixiao Zhu, Yonghui Wang,  Support vector machine learning coupled with sparse
Disease using Volumetric Chuanjun Zhuo,Qunxing Xu, Yuan principal component analysis demonstrated the best
Convolutional Neural Network and Yao, Zhuyun Liu discriminative performance, yielding classification
Transfer Learning the journal accuracies of 82.02% (AD vs. NC), 81.33% (MCI vs.
Frontiers in Aging Neuroscience NC), and 81.08% (AD vs. MCI).
Published: February 2022

2. MRI-based brain tumor image Arkapravo Chattopadhyay,

detection using CNN based deep MausumiMaitra ,  Segment brain tumors from 2D Magnetic Resonance
learning method Elsevier Journal of brain Images (MRI) by a convolutional neural network
Neuroscience Informatics 2021 which is followed by traditional classifiers and deep
learning methods.
 SVM classifier and other activation algorithms (softmax,
RMSProp, sigmoid, etc) to cross-check of work
 Implement our proposed method using “TensorFlow”
and “Keras” in “Python”

3. Predicting the Prognosis of MCI Fusun Er and Dionysis Goularas  The aim of this paper is to develop a computer-aided
Patients Using Longitudinal MRI diagnosis system with a deep-learning approach for
Data IEEE/ACM transactions on distinguishing “Mild Cognitive Impairment (MCI)
4 due to
computational biology and Alzheimer’s Disease (AD)” patients among a list of MCI
 Literature Review

S.No. Name of the Research Paper Name of the authors Summary

6. Missing MRI Pulse Sequence Sharma and G. Hamarneh.  GAN is capable of leveraging redundant information
Synthesis Using Multi-Modal IEEE Transactions on Medical contained within multiple available sequences in order to
Generative Adversarial Network” Imaging generate one or more missing sequences for a patient
2020 scan.
 The proposed network is designed to combine
information from all the available pulse sequences and
synthesizes the missing ones in a single forward Pass
7. A Novel Convolutional Variation of RUIZHI HAN 1 , C. L. PHILIP  In this paper, novel models combined with the BLS
Broad Learning System for CHEN 1,2,3, (Fellow, IEEE), AND method is presented for the diagnosis of AD, CN, and
Alzheimer’s Disease Diagnosis by ZHULIN LIU 4 Nov 2020. IEEE MCI based on MRI images
Using MRI Images TransactionsVOLUME 8, 2020

8. Applying Deep Learning to Daniel Stamate,Richard Smith, An analysis on ADNI data and determine its effectiveness
Predicting Dementia Ruslan Tsygancov, Rostislav for building classification models to differentiate the
and Mild Cognitive Impairment Vorobev, categories Cognitively Normal (CN), Mild Cognitive
John Langham, Daniel Stahl, and Impairment (MCI), and Dementia (DEM),
David Reeves IFIP( International based on tuning three Deep Learning models: two Multi-
Federation for Information Layer Perceptron (MLP1and MLP2) models 5
Processing ) Published by Springer
 Problem Formulation
1. Addressing brain diseases, a major cause of death worldwide.

2. Utilizing image analysis for predictive diagnosis.

3. Investigating image processing methods.

4. Building accurate image classification models.

5. Developing algorithms to improve classification accuracy.

6
Implementation of Proposed Technique

Figure 1: Block Diagram of of Image classification system

7
 Methods used in MRI Analysis
Method used for MRI Analysis Application Area
Prostate cancer MRI seg. Medical Imaging - Prostate Cancer
Cerebral ischemic stroke seg. Medical Imaging - Stroke
Brain tumor and stroke classification Medical Imaging - Brain Abnormalities
MRI brain tumors segmentation Medical Imaging - Brain Tumor
Radiogenomic approaches Medical Imaging - Glioma
Alzheimer's disease classification Medical Imaging - Alzheimer's Disease
Predicting Alzheimer's disease stage Medical Imaging - Alzheimer's Disease
Deuterium MR spectroscopic imaging Medical Imaging - Brain Tumor
Uncertainty-aware brain age prediction Medical Imaging - Brain Age Prediction
Multitask ensemble learning Medical Imaging - Alzheimer's Disease 8
Precision, Accuracy & Recall of different Models &
Processes

9
Results and Discussions

10
 Conclusions
• Brain disease diagnosis via image analysis is a crucial research area with significant
potential.
• Established imaging techniques like MRI, CT, PET, and SPECT have long been
instrumental in diagnosis.
• Machine learning algorithms offer promising enhancements in accuracy and
efficiency.
• Collaboration among clinicians, researchers, and machine learning experts is essential
to address challenges.
11
 Future Scope
• Enhanced Automation: Develop more automated and AI-driven diagnosis systems to
reduce human error and enhance efficiency.
• Interpretability: Focus on creating interpretable AI models to gain the trust of medical
professionals.
• Data Augmentation: Explore techniques for generating diverse and augmented datasets
to improve model robustness.
• Personalized Medicine: Tailor diagnoses and treatments based on individual patient data
and disease characteristics.
• Real-time Diagnosis: Investigate the feasibility of real-time brain disease diagnosis for
immediate intervention.
• Ethical Considerations: Address ethical and privacy concerns related to the use of
sensitive medical data.
• Integration with Healthcare Systems: Seamless integration of AI tools into clinical
workflows for practical application. 12
 References
1. P. Tupe-Waghmare et al., "Comprehensive Genomic Subtyping of Glioma Using Semi-Supervised Multi-Task Deep Learning on Multimodal MRI," in IEEE
Access, vol. 9, pp. 167900-167910, 2021, doi: 10.1109/ACCESS.2021.3136293.
2. S. Ahmad and P. K. Choudhury, "On the Performance of Deep Transfer Learning Networks for Brain Tumor Detection Using MR Images," in IEEE Access,
vol. 10, pp. 59099-59114, 2022, doi: 10.1109/ACCESS.2022.3179376.
3. Y. Zhu, X. Li, Y. Sun, H. Wang, H. Guo and J. Sui, "Investigating Neural Substrates of Individual Independence and Interdependence Orientations via
Efficiency-Based Dynamic Functional Connectivity: A Machine Learning Approach," in IEEE Transactions on Cognitive and Developmental Systems, vol.
14, no. 2, pp. 761-771, June 2022, doi: 10.1109/TCDS.2021.3101643.
4. B. Guelib, K. Zarour, H. Hermessi, B. Rayene and K. Nawres, "Same-Subject-Modalities-Interactions: A Novel Framework for MRI and PET Multi-
Modality Fusion for Alzheimer’s disease Classification," in IEEE Access, vol. 11, pp. 48715-48738, 2023, doi: 10.1109/ACCESS.2023.3276722.
5. M. Soleimani, A. Vahidi and B. Vaseghi, "Two-Dimensional Stockwell Transform and Deep Convolutional Neural Network for Multi-Class Diagnosis of
Pathological Brain," in IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 29, pp. 163-172, 2021, doi:
10.1109/TNSRE.2020.3040627.
6. P. -H. Kuo, C. -T. Huang and T. -C. Yao, "Optimized Transfer Learning Based Dementia Prediction System for Rehabilitation Therapy Planning," in IEEE
Transactions on Neural Systems and Rehabilitation Engineering, vol. 31, pp. 2047-2059, 2023, doi: 10.1109/TNSRE.2023.3267811.
7. V. Nath, D. Yang, B. A. Landman, D. Xu and H. R. Roth, "Diminishing Uncertainty Within the Training Pool: Active Learning for Medical Image
Segmentation," in IEEE Transactions on Medical Imaging, vol. 40, no. 10, pp. 2534-2547, Oct. 2021, doi: 10.1109/TMI.2020.3048055.
8. C. M. Chabib, L. J. Hadjileontiadis and A. A. Shehhi, "DeepCurvMRI: Deep Convolutional Curvelet Transform-Based MRI Approach for Early Detection of
Alzheimer’s disease," in IEEE Access, vol. 11, pp. 44650-44659, 2023, doi: 10.1109/ACCESS.2023.3272482.
9. M. S. Majib, M. M. Rahman, T. M. S. Sazzad, N. I. Khan and S. K. Dey, "VGG-SCNet: A VGG Net-Based Deep Learning Framework for Brain Tumor
Detection on MRI Images," in IEEE Access, vol. 9, pp. 116942-116952, 2021, doi: 10.1109/ACCESS.2021.3105874.
10.R. Haweel et al., "A Novel Grading System for Autism Severity Level Using Task-Based Functional MRI: A Response to Speech Study," in IEEE Access,
vol. 9, pp. 100570-100582, 2021, doi: 10.1109/ACCESS.2021.3097606.
11.Dinesh Bhoyar, Barkha Katey, Manish Ingle,” LoRa Technology Based Low Cost Water Meter Reading System, Proceedings of 3rd International
Conference on Internet of Things and Connected Technologies (ICIoTCT) 26-27 2018/5/3 13