Resource-Efficient Medical Report Generation using Large Language Models

TL;DR

This paper presents a resource-efficient framework using vision-enabled Large Language Models to automatically generate accurate and contextually rich radiology reports from chest X-ray images, aiming to assist radiologists and promote clinical automation.

Contribution

It introduces a lightweight, vision-enabled LLM framework with optimization techniques like prefix tuning for medical report generation, achieving competitive performance on MIMIC-CXR dataset.

Findings

Outperforms previous methods in report quality and efficiency

Achieves strong medical contextual understanding

Demonstrates resource efficiency and high precision

Abstract

Medical report generation is the task of automatically writing radiology reports for chest X-ray images. Manually composing these reports is a time-consuming process that is also prone to human errors. Generating medical reports can therefore help reduce the burden on radiologists. In other words, we can promote greater clinical automation in the medical domain. In this work, we propose a new framework leveraging vision-enabled Large Language Models (LLM) for the task of medical report generation. We introduce a lightweight solution that achieves better or comparative performance as compared to previous solutions on the task of medical report generation. We conduct extensive experiments exploring different model sizes and enhancement approaches, such as prefix tuning to improve the text generation abilities of the LLMs. We evaluate our approach on a prominent large-scale radiology report dataset - MIMIC-CXR. Our results demonstrate the capability of our resource-efficient framework to generate patient-specific reports with strong medical contextual understanding and high precision.