MUMBAI, India, Oct. 5 -- Intellectual Property India has published a patent application (202611100777 A) filed by Swami Rama Himalayan University on August 20, 2026, for A System And Method For Fractal-Guided Explainable Hybrid Ai Skin Lesion Classification.
Inventors include Dr. Suman Pant; Prerit; Deepesh Solanki; Priyanshu Singh; and Aman Panwar.
The application for the patent was published on October 02, 2026, under issue no. 40/2026.
Abstract: The invention relates to explainable skin lesion classification in data-scarce clinical environments. A skin lesion image from clinical or remote imaging sources is preprocessed and segmented to isolate a lesion region. A Fractal Complexity Vector is then generated using Boundary Fractal Dimension, Differential Box-Counting Fractal Dimension, lacunarity analysis, multi-fractal spectrum analysis, asymmetry evaluation, and structural irregularity indicators. The fractal-derived descriptors, and optionally local fractal response maps, are transformed into spatial importance weights that guide a Hybrid AI Learning Module during deep feature extraction. A Hybrid Fractal-Feature Fusion Framework combines a fractal descriptor vector and a deep neural feature vector to classify lesion category and malignancy probability. Weighted attribution fusion produces Fractal Attribution Explanation Maps, and a Dual Reliability Estimation Mechanism generates a confidence score from an Irregularity Consistency Score and a Predictive Uncertainty Score. Score, enabling interpretable diagnostic outputs for physicians.
Disclaimer: Curated by HT Syndication.
