MUMBAI, India, Jan. 2 -- Intellectual Property India has published a patent application (202541123218 A) filed by Vellore Institute Of Technology, Vellore, Tamil Nadu, on Dec. 6, 2025, for 'vision language model for medical image diagnosis with attention-based explainability.'
Inventor(s) include Dr. S. Ebenezer Juliet; Mr. Abilash R; and Mr. Jermish Justin.
The application for the patent was published on Jan. 2, under issue no. 01/2026.
According to the abstract released by the Intellectual Property India: "The present disclosure provides a system for automated medical image analysis. The system includes a vision-language model configured to receive medical images and generate textual captions describing medical findings, an attention visualization module configured to extract attention information and generate visual representations highlighting image regions contributing to caption generation, a medical term extraction module configured to identify medical terminology from generated captions, a reference retrieval module configured to retrieve supporting medical literature based on extracted medical terms, and an output module configured to combine the textual caption, visual representations, and retrieved literature into integrated output. The vision-language model incorporates bootstrapping language-image pretraining with low-rank adaptation fine-tuning. The attention visualization module extracts cross-attention weights and generates heatmaps overlaid on medical images. The medical term extraction module filters terms based on predefined medical entity types."
Disclaimer: Curated by HT Syndication.
