Alzheimer's Disease Detection using VGG19 Model on MRI Scans

The application of Convolutional Neural Networks (CNNs) in detecting Alzheimer's Disease from MRI scans is presented by the authors in this work. The study investigates the use of SMOTE (Synthetic Minority Over-Sampling Technique) to deal with class imbalance inherent in medical datasets. A comprehensive evaluation of CNN models trained on balanced datasets using SMOTE is presented, highlighting their effectiveness in improving diagnostic accuracy. It explores the use of the Inception CNN architecture for detecting Alzheimer's disease (AD) from MRIs. We leverage transfer learning by utilizing a pretrained InceptionV3 model, fine-tuned for the classification task withsignificant accuracy of 96% is obtained and shown in result section.