MUMBAI, India, Aug. 10 -- Intellectual Property India has published a patent application (202621073876 A) filed by Symbiosis International Deemed University on June 12, 2026, for A System And Method For Brain Age Prediction Using Structural Mri And Deep Learning.
Inventors include Dr. Princy Diwan; Ritika Mohan; Yukti Madan; and Himangii Chatterjee.
The application for the patent was published on July 31, 2026, under issue no. 31/2026.
Abstract: ABSTRACT A SYSTEM AND METHOD FOR BRAIN AGE PREDICTION USING STRUCTURAL MRI AND DEEP LEARNING The present invention provides a deep learning based system (100) and method for predicting biological brain age from structural magnetic resonance imaging (MRI) data for early neurodegenerative risk assessment. The system (100) comprises a data acquisition and preprocessing module (110) that extracts mid-axial two-dimensional slices from three-dimensional NIfTI format MRI volumes and performs intensity normalization and channel conversion. A deep learning prediction module (120) employs a pre-trained ResNet convolutional neural network architecture adapted through transfer learning, with a modified regression head (123) for predicting brain age as a single continuous value. An evaluation and risk assessment module (130) computes mean absolute error, classification accuracy, F1-score, and brain age gap as quantitative biomarkers. A visualization module (140) generates training loss convergence curves and prediction error distribution charts. The system uniquely combines two-dimensional slice- based transfer learning with a comprehensive multi-metric evaluation framework for clinically viable neurodegenerative risk assessment. [
Disclaimer: Curated by HT Syndication.
