This project was conducted as an Introduction to implementing CNN for classification of Alzheimers Progression using MRI scans. An exploratory study was conducted to learn how to use deep learning for feature extraction and training a machine learning model on the features extracted for dementia classification. The experiment was conducted for Binary classification (Dementia v/s No Dementia) and Multiclass Classification (Very Mild Dementia, Mild Dementia, Moderate Dementia, No Dementia).
The dataset utilized for this study can be found at this link - https://www.kaggle.com/datasets/tourist55/alzheimers-dataset-4-class-of-images?datasetId=457093&sortBy=voteCount
The Deep Learning model achieves over 95% recall on the validation dataset. But performs poorly on the testing dataset, with a maximum recall of 67%.