Deep Learning for Alzheimer's Diagnosis: ResNet152V2 Approach on MRI Dataset

Abstract

Alzheimer's Disease (hereafter AD), a progressive neurodegenerative disorder, poses a significant global health challenge. This research presents a convolutional neural network (CNN)-based algorithm utilizing the ResNet152V2 architecture to classify AD severity from MRI images. Our method makes use of machine learning to reliably identify the various stages of AD, allowing for an early and accurate diagnosis. The ResNet152V2 architecture demonstrates exceptional performance, achieving a test accuracy of 97.31%. The model's robustness is validated through comprehensive evaluations, including precision, recall, and f1-score This research contributes to the field by providing a reliable and effective tool for the automated detection of AD severity, potentially aiding clinicians in timely intervention and treatment planning. The utilization of augmented MRI data enhances the model's generalization capability, making it adaptable to diverse clinical scenarios. � 2024 IEEE.