Alzheimer’s Disease Diagnosis from MRI images using ResNet-152 Neural Network Architecture

Alzheimer’s disease is a degenerative brain illness that gradually weakens memory and thinking abilities, as well as the capacity to do even the most basic tasks. Classification strategies are required to figure out the stages of this disease. In this paper, we have proposed a technique to differentiate Alzheimer’s brain image from a healthy brain image using Convolutional Neural Network (CNN). We have also done the work of multiclass classification for Alzheimer’s disease diagnosis. Classification of different stages of this disease using clinical data has been challenging and problematic as the extraction of discriminative and selective features from the medical images such as MRI, fMRI is hard to find. We have used the ResNet-152 architecture to diagnose Alzheimer’s disease successfully where we classified not only two stages of this disease to reach test accuracy of 99.30% but also four stages of this disease to reach test accuracy of 98.79%. Our model outperforms many of the existing methods in terms of accuracy.