Detection of Alzheimer’s disease from temporal lobe grey matter slices using 3D CNN

ABSTRACT

Magnetic Resonance Imaging helps in detecting brain atrophy caused by Alzheimer’s disease (AD). Morphological changes start to occur in the temporal lobe of the brain in persons affected by AD. A 3D Convolutional Neural Network is proposed that focuses on the temporal lobe of the brain for AD classification from normal controls. By investigating the temporal region of the grey matter in the brain instead of the entire brain region high performance is obtained with an accuracy of 88.06%, a sensitivity of 94.03%, and a specificity of 82.09% in classifying the MRI volumes of Alzheimer’s Disease Neuroimaging Initiative (ADNI-2) dataset. When Apolipoprotein ε4 allele information and mini-mental state examination scores are used with the MRI volumes, an accuracy of 89.55%, a sensitivity of 82.09% and a specificity of 97.01% for ADNI-2 MRI volumes are achieved. When tested on ADNI-3 MRI volumes the proposed method resulted in an accuracy of 90.14%, a sensitivity of 79.17% and a specificity of 92.37%.

KEYWORDS:

• Magnetic resonance imaging
• Alzheimer’s disease
• 3D convolutional neural network
• grey matter
• temporal lobe
• Apolipoprotein
• Mini-Mental State Examination
• Brain Atrophy

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Data used in the current study were collected by ADNI and downloaded through the LONI platform after being approved by the data access committee.

Additional information

Funding

Data collection and sharing for this project was funded by the Alzheimer’s Disease Neuroimaging Initiative (ADNI) (National Institutes of Health Grant U01 AG024904) and DOD ADNI (Department of Defense award number W81XWH-12-2-0012). ADNI data are disseminated by the Laboratory for Neuro Imaging at the University of Southern California.

Notes on contributors

R. Divya

R. Divya has finished her B.E. in Electronics and Communication Engineering, and M.E. in Embedded and Real Time Systems, at PSG College of Technology, Coimbatore, India. Currently, she is doing her Ph.D. at Mepco Schlenk Engineering College, Sivakasi, India. She has 2 years of teaching and research experience in the field of Model-Based Testing and Machine Learning. She has published 3 research papers in reputed international journals. She is a life member of ISTE.

R. Shantha Selva Kumari

Dr. R. Shantha Selva Kumari has finished her B.E. in Electronics and Communication Engineering, M.S. in Electronics & Control, and Ph.D. in Bio Signal Processing. Currently, she is working as the Senior Professor and Head in the Department of Electronics and Communication Engineering, Mepco Schlenk Engineering College, Sivakasi, India. She has 33 years of teaching and research experience in the field of Bio Signal Processing and Digital Communication. She has published more than 100 research papers in reputed international journals and more than 150 papers at various conferences. She has filed 3 patents and was granted 5 copyrights. She is a life member of FIETE, ISTE, and CSI.