MUMBAI, India, July 30 -- Intellectual Property India has published a patent application (202641088192 A) filed by Kovvuru Guru Joshna Devi; and Vamshi Krishna B on July 20, 2026, for A System And Method For Explainable Skin Lesion Prediction And Clinical Decision Support Using Hybrid Deep Learning And Multi-Modal Dermoscopic Image Analysis.
Inventors include Kovvuru Guru Joshna Devi; and Vamshi Krishna B.
The application for the patent was published on July 24, 2026, under issue no. 30/2026.
Abstract: The present invention discloses a system and method for explainable skin lesion prediction and clinical decision support using hybrid deep learning and multi-modal dermoscopic image analysis. The invention integrates dermoscopic images, patient demographic information, lesion metadata, and clinical observations through adaptive multi-modal feature fusion. Multiple deep learning architectures including convolutional neural networks, transformer-based attention mechanisms, and explainable artificial intelligence techniques generate accurate and interpretable diagnostic predictions. The framework produces Grad-CAM heatmaps, SHAP feature importance scores, lesion localization maps, confidence estimation, and textual diagnostic explanations to improve clinician trust. A clinical decision support module performs malignancy risk assessment, lesion severity estimation, treatment recommendation assistance, and referral prioritization. The proposed invention significantly enhances diagnostic accuracy, transparency, reliability, and clinical decision-making while supporting deployment in hospitals, dermatology clinics, and telemedicine environments.
Disclaimer: Curated by HT Syndication.
