MUMBAI, India, Aug. 24 -- Intellectual Property India has published a patent application (202621082033 A) filed by Symbiosis International Deemed University on July 02, 2026, for Self-Modifying Nested Learning System And Method For Kidney Stone Classification.
Inventors include Dr. Aniket K. Shahade; and Priyanka V. Deshmukh.
The application for the patent was published on August 21, 2026, under issue no. 34/2026.
Abstract: ABSTRACT SELF-MODIFYING NESTED LEARNING SYSTEM AND METHOD FOR KIDNEY STONE CLASSIFICATION A self-modifying nested learning system (100) and method for kidney stone classification from axial CT images is disclosed. The system (100) comprises an input module (110), a feature extraction backbone (120) employing a pretrained ResNet-18 architecture, a deep surprise predictor module (130) generating adaptive modulation signals from exponential moving average statistics, a fast-weight memory module (140) that continuously updates a fast-weight matrix through a Hebbian-like outer product-based learning rule during inference, a multi-timescale nested learning module (150) with temporally differentiated slow-learning layers, and a classification head (160). The system (100) emulates biological synaptic plasticity, enabling dynamic adaptation of internal representations during inference. The system (100) achieves 95.1 percent classification accuracy with 94.8 percent F1-score, outperforming static deep learning baselines while adding only approximately 15 percent computational overhead. [
Disclaimer: Curated by HT Syndication.
