MUMBAI, India, April 17 -- Intellectual Property India has published a patent application (202621017407 A) filed by Prof. Shubhangi Kamble; Saloni Tiwari; Darshan Shete; Sairaj Patil; and Sara Ansingkar, Pune, Maharashtra, on Feb. 17, for 'report generation using meta-classifier.'
Inventor(s) include Prof. Shubhangi Kamble; Saloni Tiwari; Darshan Shete; Sairaj Patil; and Sara Ansingkar.
The application for the patent was published on April 17, under issue no. 16/2026.
According to the abstract released by the Intellectual Property India: "ReportGen is a novel AI-based tool for the automated generation of radiology reports that addresses the weaknesses of single-model techniques by utilizing a meta-classifier model. ReportGen can automatically train three different models based on CNN, Attention, and Region concepts, each of them capable of generating a candidate radiology report from the same chest X-ray image. Instead of choosing only one model, ReportGen selfevaluates all the candidate reports based on confidence scores, linguistic qualities (BLEU, ROUGE, BERTScore), as well as the agreement of clinical labels with CheXpert labels. Then, a meta-classifier model can predict and determine the best and trustworthy clinical report. The experiment conducted using the Indiana University Chest X-ray dataset will validate the efficiency and improved clinical accuracy of the more robust approach by comparing the outcome with other single models. ReportGen will provide a significant contribution towards the applicability of AI by avoiding inconsistencies between various types of clinical cases and minimizing the jeopardy of diagnostic errors."
Disclaimer: Curated by HT Syndication.
