| MS Seminar

Name of the Speaker: Ms. Bhumika Gupta (EE21S060)
Guide: Dr. Mohanasankar S
Online meeting link: https://meet.google.com/mdu-kuyr-aht
Date/Time: 23rd April 2025 (Wednesday), 4:00PM
Title: Comprehensive Knowledge Modeling and Question Answering of Cancer Clinical Practice Guidelines using Large Language Models.

Abstract :

Clinical Practice Guidelines (CPGs) are crucial tools for healthcare professionals, often presented as graphical flows to outline diagnostic and treatment pathways. This research presents an innovative approach to enhance the utility of Clinical Practice Guidelines (CPGs) by developing a contextually enriched digital representation of cancer guidelines. By leveraging Large Language Models (LLMs) for the automated extraction and classification of nodes and relationships within CPGs, this work significantly improves the accuracy of knowledge modeling. The study also introduces a novel method for answering natural language questions by extracting and utilizing relevant subgraphs from the structured CPG knowledge model, ensuring factual accuracy in medical question answering. Additionally, the research explores the use of Multimodal-LLMs (MLLMs) for direct question answering from unstructured CPG documents, offering a comparative analysis of both structured and unstructured approaches.