Applying stacked sparse autoencoders and artificial neural networks for the prediction of biological age of the brain using images from magnetic resonance imaging

• Grant number: 19/08553-5
• Support type: Scholarships in Brazil - Scientific Initiation
• Effective date (Start): August 01, 2019
• Effective date (End): July 31, 2020
• Field of knowledge: Physical Sciences and Mathematics - Computer Science
• Principal researcher: Ricardo José Ferrari
• Grantee: Paulo Henrique Dal Bello
• Home Institution: Centro de Ciências Exatas e de Tecnologia (CCET). Universidade Federal de São Carlos (UFSCAR). São Carlos , SP, Brazil

Abstract

Alzheimer's disease (AD) is the most common form of dementia, accounting for about 60 to 80% of all cases and affecting mainly individuals over age 65. In Brazil, about one million people have already been affected by AD. Although there is no cure for AD at the moment, the early diagnosis of this disease allows the patient to undergo treatments that delay or even revert cognitive loss. It is well known that normal brain aging follows a specific pattern, which allows the prediction of brain age based on its degeneration. This evaluation allows the early prediction of neurocognitive disorders, such as AD. Based on the foregoing, this research has the objective of developing an artificial neural network (ANN) to classify biological age out of cerebral anatomical characteristics extracted from hippocampal magnetic resonance imaging scans via stacked sparse autoencoders. The ANN will be developed using only images of cognitively normal patients and will subsequently be applied to evidence AD in patients that, by any chance, show much more advanced cerebral degeneration than the expected for their age.