MedEval: A Multi-Level, Multi-Task, and Multi-Domain Medical Benchmark for Language Model Evaluation

TL;DR

MedEval is a comprehensive, multi-level, multi-task, multi-domain medical benchmark designed to evaluate and improve language models in healthcare, highlighting the importance of instruction tuning for effective medical language understanding.

Contribution

This paper introduces MedEval, a novel medical benchmark with extensive annotated datasets across multiple domains and tasks, enabling systematic evaluation of language models in healthcare.

Findings

Large language models show variable effectiveness across tasks.

Instruction tuning enhances few-shot performance of large models.

Benchmarking reveals strengths and limitations of models in medical contexts.

Abstract

Curated datasets for healthcare are often limited due to the need of human annotations from experts. In this paper, we present MedEval, a multi-level, multi-task, and multi-domain medical benchmark to facilitate the development of language models for healthcare. MedEval is comprehensive and consists of data from several healthcare systems and spans 35 human body regions from 8 examination modalities. With 22,779 collected sentences and 21,228 reports, we provide expert annotations at multiple levels, offering a granular potential usage of the data and supporting a wide range of tasks. Moreover, we systematically evaluated 10 generic and domain-specific language models under zero-shot and finetuning settings, from domain-adapted baselines in healthcare to general-purposed state-of-the-art large language models (e.g., ChatGPT). Our evaluations reveal varying effectiveness of the two categories of language models across different tasks, from which we notice the importance of instruction tuning for few-shot usage of large language models. Our investigation paves the way toward benchmarking language models for healthcare and provides valuable insights into the strengths and limitations of adopting large language models in medical domains, informing their practical applications and future advancements.