Early Prediction of Alzheimer Disease and Multiclass Classifier System

Alzheimer disease (AD) is a neurological disorder damaging the memory systems of the brain that gets worse as time passes and significantly affects neuropsychological function.These days, the field of medical imaging is significantly evolving due to advancements in deep learning (DL). Large datasets can be effectively employed using DL, such as the Alzheimer Disease dataset, which has several classifications. In this project, 6400 MRI images that have been reduced to 256 × 256 pixels and are separated into four unique categories are used to create a RBF kernel based SVM model for multistage classification and detection of Alzheimer's MRI pre-processed dataset. The dataset with 4 classes was used for this detection and was used for the purpose of training and testing our SVM model, giving the detection accuracy 86% and classification accuracy 78%. The resulting model implements the automated AD detection and staging using MRIs of brain, providing the valuable tool for clinicians in diagnosing and monitoring the development of Alzheimer disease.