MedGellan: LLM-Generated Medical Guidance to Support Physicians

TL;DR

MedGellan is a lightweight framework leveraging LLMs to generate clinical guidance from raw medical records, aiding physicians in diagnosis with improved recall and F1 scores, while respecting data temporal order.

Contribution

It introduces a novel Bayesian-inspired prompting strategy for LLMs to generate medical guidance without annotations, enhancing diagnostic support.

Findings

Improved diagnostic recall and F1 score with MedGellan guidance

Guidance generation respects temporal order of clinical data

Framework operates without requiring annotated data

Abstract

Medical decision-making is a critical task, where errors can result in serious, potentially life-threatening consequences. While full automation remains challenging, hybrid frameworks that combine machine intelligence with human oversight offer a practical alternative. In this paper, we present MedGellan, a lightweight, annotation-free framework that uses a Large Language Model (LLM) to generate clinical guidance from raw medical records, which is then used by a physician to predict diagnoses. MedGellan uses a Bayesian-inspired prompting strategy that respects the temporal order of clinical data. Preliminary experiments show that the guidance generated by the LLM with MedGellan improves diagnostic performance, particularly in recall and $F_1$ score.