Brain Age Estimation based on Brain MRI by an Ensemble of Deep Networks

Abstract

Estimation of biological brain age is one of the topics that has been much discussed in recent years. One of the most important reasons for this is the possibility of early detection of neurodegenerative disorders such as Alzheimer's and Parkinson's with Brain Age Estimation (BAE). Brain imaging is one of the most important data to estimate the biological age of the brain. Because the brain's natural aging follows a particular pattern, it enables researchers and physicians to predict the human brain's age from its degeneration. Some studies have been done on 2D or 3D brain images data for this purpose. In this study, an ensemble structure, including 3D and 2D Convolutional Neural Networks (CNNs), is used to BAE. The proposed ensemble CNN (ECNN) method obtained a Mean Absolute Error (MAE) of 3.57 years, which is better than the previous studies.