Conference proceeding
Beyond Self-consistency: Ensemble Reasoning Boosts Consistency and Accuracy of LLMs in Cancer Staging
ARTIFICIAL INTELLIGENCE IN MEDICINE, PT I, AIME 2024, v 14844, pp 224-228
01 Jan 2024
Abstract
Pathologic cancer stage, crucial for treatment decisions, is often buried in unstructured pathology reports. This study investigates using pre-trained clinical LLMs for stage extraction, leveraging prompting techniques like chain-of-thought to enhance model transparency. While self-consistency methods further improve LLM performance, they can introduce inconsistencies in reasoning paths and predictions. We propose an ensemble reasoning approach, aiming for reliable cancer stage extraction. Utilizing an open-source clinical LLM on real-world reports, we demonstrate that the ensemble approach improves consistency and boosts performance, paving the way for utilizing LLMs in healthcare settings where reliability and interpretability are paramount.
Metrics
1 Record Views
Details
- Title
- Beyond Self-consistency: Ensemble Reasoning Boosts Consistency and Accuracy of LLMs in Cancer Staging
- Creators
- Chia-Hsuan Chang - Drexel UniversityMary M. Lucas - Drexel UniversityYeawon Lee - Drexel UniversityChristopher C. Yang - Drexel UniversityGrace Lu-Yao - Thomas Jefferson University
- Contributors
- J Finkelstein (Editor)R Moskovitch (Editor)E Parimbelli (Editor)
- Publication Details
- ARTIFICIAL INTELLIGENCE IN MEDICINE, PT I, AIME 2024, v 14844, pp 224-228
- Series
- Lecture Notes in Artificial Intelligence
- Publisher
- Springer Nature
- Number of pages
- 5
- Grant note
- National Science Foundation; National Science Foundation (NSF) DoD W91XWH-05-1-023 / Department of Defense; United States Department of Defense
- Resource Type
- Conference proceeding
- Language
- English
- Academic Unit
- Information Science
- Web of Science ID
- WOS:001295129500023
- Scopus ID
- 2-s2.0-85200751905
- Other Identifier
- 991022202101004721