Performance Analysis of Machine Learning and Deep Learning Models for Classification of Alzheimer's Disease from Brain MRI

Abstract

Alzheimer's disease (AD) is an irreversible and degenerative brain condition that gradually damages memory and thinking abilities. Despite being incurable, AD causes significant pain and financial hardship to patients and their families. However, medications are most effective when administered early in the course of the disease and early diagnosis is crucial in the treatment of AD to restrict its progression. There are several approaches proposed for computer-assisted AD diagnosis that involve structural and functional imaging modalities, such as sMRI, fMRI, DTI, and PET. Machine learning and deep learning techniques have facilitated the development of novel models for diagnostic accuracy in AD. This research compares the performance of several machine learning and deep convolutional architectures to detect AD from MCI. It is essential to find the effective baseline model for classifying AD, hence all the pre-trained models are evaluated with benchmark dataset. Experimental observations indicate that the DenseNet-169 performed best out of different state-of-the-art architectures, with an average accuracy of 82.2%.