Detecting Alzheimer's disease through analysis of MRI images using deep learning

Abstract - This paper focuses on developing a deep learning framework to detect different stages of Alzheimer's disease using MRI images. The ADNI dataset, which categorizes subjects into four stages — Non-demented, Very Mild-Demented, Mild- Demented, and Moderate-Demented—serves as the foundation. These stages range from healthy individuals to those exhibiting varying levels of cognitive decline. The work aims to accurately classify the stage of Alzheimer's based on MRI scans, providing a tool for early detection and monitoring of the disease. The scope of the work includes data collection, preprocessing, feature extraction, and the application of convolutional neural networks (CNNs) to identify key patterns in brain imaging data. The scope of the project includes data collection, preprocessing, feature extraction, and the application of convolutional neural networks (CNNs) here we have applied 5 CNN models namely Convo 2D, ResNet50, VGG16, Inception, Mobile Net to identify key patterns in brain imaging data. By leveraging CNNs, the model is able to capture intricate features associated with brain degeneration. The process involves model selection, training, and fine-tuning to enhance performance, followed by comprehensive evaluation to ensure accuracy across different Alzheimer's stages. Results show that the deep learning model effectively distinguishes between the four stages of Alzheimer's, achieving high classification accuracy. Among the implemented CNN models Convo 2D is showing highest accuracy of 99%.