Automated detection of Alzheimer’s Disease using Deep Learning in MRI

• Published in: Journal of physics. Conference series Vol. 1921; no. 1; pp. 12024 - 12031
• Main Authors: Muhammed Raees, P C, Thomas, Vinu
• Format: Journal Article
• Language: English
• Published: Bristol IOP Publishing 01.05.2021
• Subjects: Algorithms; Alzheimer's disease; Artificial neural networks; Automation; Deep learning; Degeneration; Image classification; Machine learning; MRI; Physics; RELU; Software; Support vector machines; Transfer learning
• ISSN: 1742-6588; 1742-6596; 1742-6596
• DOI: 10.1088/1742-6596/1921/1/012024

Alzheimer’s disease (AD) is a progressive mental deterioration and incurable neurodegenerative disease that can occur in middle or old age, due to generalized degeneration of the brain. Because of the irreversible nature of the progression of Alzheimer’s disease, the early diagnosis of AD has an immense clinical, social, and economic need. This research output proposing a state-of-the-art, easy, and early automated deep learning-based system to predict AD from a large MRI dataset of normal and diseased subjects. It classified the database of 111 subjects into Mild Cognitive Impairment (MCI), Alzheimer’s disease (AD), and Normal classes. Classification tools like Support Vector Machines (SVM) and different models of Deep Neural Network (DNN) algorithms were tested. Deep learning algorithms were offering high accuracy of about 80-90% on AD prediction. For the prediction of diseases such as Alzheimer’s, Dementia, and Parkinson’s, applying highly accurate computational-automated machine learning-tools will help to diagnose the disease in the early stage itself and provide a better clinical, social and economic outcome.