MUMBAI, India, Aug. 17 -- Intellectual Property India has published a patent application (202641096817 A) filed by Sanjiv Kumar; Dr. Ankur Dumka; and Dr. Sushil Kumar on August 11, 2026, for A System And Method For Multi-Objective Evolutionary Hyperparameter Optimization Of Deep Learning Models For Brain Tumor Segmentation.
Inventors include Sanjiv Kumar; Dr. Ankur Dumka; and Dr. Sushil Kumar.
The application for the patent was published on August 14, 2026, under issue no. 33/2026.
Abstract: The present invention discloses a system and method for multi-objective evolutionary hyperparameter optimization of deep learning models for brain tumor segmentation. The system receives brain MRI images and corresponding segmentation masks and performs preprocessing to generate standardized image representations. An evolutionary optimization module generates candidate hyperparameter configurations for a deep learning segmentation model and evaluates each configuration using multiple segmentation objectives comprising accuracy, sensitivity, specificity, and Dice similarity coefficient. The evolutionary optimization module applies selection and evolutionary operations to iteratively improve the candidate configurations. Pareto-based optimization is employed to identify non- dominated hyperparameter configurations representing different performance trade-offs. An optimized configuration is selected from the Pareto-optimal solution set and used for training a deep learning segmentation model. The trained model processes brain MRI images and generates a segmentation map identifying tumor regions. The invention provides automated multi-objective hyperparameter optimization, reduces dependency on manual parameter selection, and facilitates balanced optimization of multiple performance measures for brain tumor segmentation.
Disclaimer: Curated by HT Syndication.
