MUMBAI, India, July 24 -- Intellectual Property India has published a patent application (202641078391 A) filed by Sri Manakula Vinayagar Engineering College on June 25, 2026, for Deep Learning-Based Multimodal Skin Cancer Detection System Using Real- Time Image Capture And Patien.
Inventors include S. Suguna; T. Logasundari; Aravind; Dr. A. V. Srinath; and Ramya D.
The application for the patent was published on July 17, 2026, under issue no. 29/2026.
Abstract: The titled invention discloses a deep learning-based multimodal skin cancer detection system that combines dermoscopic image analysis and patient metadata to improve the accuracy and reliability of skin cancer diagnosis. The Pi Camera (2) captures 10 dermoscopic skin lesion images, while patient information such as age, gender, and lesion location is collected through the Metadata Entry Module (8). The captured images are processed by the Image Preprocessing Module (7) and analyzed using the CNN Model (11) to extract diagnostic image features. Simultaneously, patient metadata is processed by the Metadata Preprocessing Module (9), and both image 15 and metadata features are integrated through the Feature Fusion Layer (10) to create a comprehensive multimodal representation. The fused data is analyzed by the Prediction Layer (12) to classify lesions as benign or malignant with associated confidence scores. The system further incorporates explainable artificial intelligence through Grad-CAM visualization to highlight lesion regions influencing the prediction, 20 thereby improving transparency and clinical trust. The entire framework is deployed on the Raspberry Pi 4 Model (5), enabling portable, real-time, and offline operation suitable for rural and low-resource healthcare environments. Experimental evaluation demonstrates high diagnostic performance with accuracy exceeding 95%, an F1- score of 94.5%, and an AUC of 0.97. The invention provides a cost-effective, 25 interpretable, and Al-assisted skin cancer screening solution that supports early detection, reduces diagnostic workload, and enhances accessibility to dermatological care.
Disclaimer: Curated by HT Syndication.
