A Study of Computer-Assisted Diagnosis of Alzheimer's Disease by Deep Learning Models

This chapter presents a study of computer-assisted diagnosis (CAD) of Alzheimer’s disease by utilizing deep learning models. Alzheimer’s Disease (AD) is an incurable, progressively degenerative neurological disorder that impairs the human brain. If it is not treated in a timely manner, it may accelerate the progression of dementia, resulting in the deterioration of memory and cognitive abilities, and eventually affect daily functioning. Manual detection of AD from magnetic resonance imaging (MRI) scans is a time-consuming and difficult task for clinicians and often fraught with low efficiency. Consequently, there is a pressing demand for an automated CAD system that is capable of detecting AD in its early stages with enhanced efficacy. In this study, we have investigated numerous methodologies of CAD, with a primary focus on MRI-based approaches. We have discussed several deep learning architectures and analyzed their performance on imbalanced MRI datasets such as the “OASIS dataset”. In particular, we have explored the deep learning models, such as convolutional neural networks and Vision Transformer models, proposed by several researchers for detecting Alzheimer’s disease.