Detection of Alzheimer’s Disease from Brain MRI Images Using Convolutional Neural Network

Alzheimer's disease (AD) ranks as the predominant factor leading to the onset of dementia. AD leads to a steady decline in memory, reasoning, behavior, and social abilities. These changes affect a person's personality. Alzheimer's disease is most common in older adults. Conventional diagnostics methods for Alzheimer's disease face challenges like invasiveness, cost, imprecision, and patient discomfort. The aim is to develop an innovative AI model offering a non-invasive, cost-effective solution. In this suggested system, we use a convolutional neural network to classify AD into four classes: Normal Phase, First Phase, Second Phase, and Third Phase. The dataset size is 6000 MRI (Magnetic Resonance Imaging) images. AD causes the brain regions to shrink, and connections between the network of neurons may break down. Hence, accurate and timely Alzheimer's disease classification is crucial for effective treatment and better patient outcomes. The primary goal of the suggested model is to classify AD with greater accuracy using a deep learning model. The work follows several preprocessing steps for the input images, such as image resizing, image pixel resizing, and batching images to enhance the input features, and for disease classification, the Convolutional Neural Network model is used. The suggested model was evaluated using Kaggle dataset images, and a maximum accuracy of 97.25% was obtained using the Kaggle dataset. This research addresses the critical need for precise AD classification, potentially revolutionizing patient outcomes through early detection.