Alzheimer’s Disease Diagnosis and Categorization using Deep Learning

Alzheimer's disease (AD) is a significant global health concern, accounting for the sixth highest cause of mortality worldwide. Early diagnosis is critical to improving patient care and treatment outcomes. This study aims to classify Alzheimer's disease (AD) using Magnetic Resonance Imaging (MRI) brain images, allowing for prompt intervention and optimal disease treatment. The dataset consists of 6400 MRI pictures classified into four categories: very mildly demented, mildly demented, moderately demented, and non-demented, with an 80-10-10 split for training, validation, and testing. Initially, the utilization of VGG16 architecture results with an accuracy of 87%. To improve accuracy and reduce training time, the GoogleNet (Inception V3) Convolutional Neural Network (CNN) architecture was used. This method resulted with an accuracy of 97.62%.