CLASSIFICATION OF ALZHEIMER'S DISEASE STAGES THROUGH VOLUMETRIC ANALYSIS OF MRI DATA

Neurodegenerative disorders are among the most important problems to be addressed in nations with increasingly aging populations. As a result of these illnesses, neurons in the human brain begin to deteriorate and eventually die, impairing movement or mental function. One of the most talked-about neurological disorders is Alzheimer's disease (AD). Finding the condition as soon as possible is critical as the best chance for survival with Alzheimer's is an early diagnosis. Magnetic resonance imaging (MRI) is one method for diagnosing AD since it may reveal structural abnormalities in the brain. Modern computer vision is highly suited for this image-based categorization challenge. Convolutional neural networks that have been trained expressly for machine learning, are what we suggest can help with early detection. The primary method of diagnosis is the segmentation of the hippocampus region of MRI images. Our approach implements a deep learning model that is trained on an MRI dataset. Using the volume of the hippocampus, this method correctly divides the MRI scans into the four phases of AD i.e. Mildly Demented, Moderately Demented, Non-Demented, and Severely Demented.