Brain age prediction from MRI images based on a convolutional neural network with MRMR feature selection layer

An sophisticated medical technique used to diagnose illnesses and brain disorders including multiple sclerosis, Alzheimer's, and other neurological ailments is the ability to predict the biological age of the brain using MRI pictures. To do this, sophisticated algorithms and neural networks are used to scan MRI brain pictures in order to extract different brain properties, including cortical thickness and volume. The brain ages of individuals are determined by matching their characteristics against MRI imaging data collected from other patients. The research employs a new deep learning model named CNN-MRMR which combines features from the Minimum Redundancy Maximum Relevance (MRMR) feature selection approach and Convolutional Neural Network (CNN) technology. MRI images of human brains are initially processed by the convolutional network to extract age-related characteristics. The feature selection layer uses MRMR algorithm which identifies essential characteristics for a target variable while minimizing feature redundancy to select the optimal feature subset. The system employs a regression layer as the final stage to predict brain age by utilizing the selected characteristics. The proposed method for estimating individual brain age attained a prediction accuracy of 90.3%, outperforming results from comparable research studies.