ERVQA: A Dataset to Benchmark the Readiness of Large Vision Language Models in Hospital Environments

TL;DR

This paper introduces ERVQA, a new dataset and benchmark for evaluating large vision-language models' ability to handle emergency room scenarios in hospitals, revealing their current limitations and the need for domain-specific improvements.

Contribution

The paper presents ERVQA, the first comprehensive dataset and benchmark for assessing LVLMs in hospital emergency scenarios, including detailed error analysis and evaluation metrics.

Findings

ERVQA is a highly complex task highlighting current model limitations.

Analysis shows model performance varies with decoder type and size.

Domain-specific solutions are necessary for improved hospital environment understanding.

Abstract

The global shortage of healthcare workers has demanded the development of smart healthcare assistants, which can help monitor and alert healthcare workers when necessary. We examine the healthcare knowledge of existing Large Vision Language Models (LVLMs) via the Visual Question Answering (VQA) task in hospital settings through expert annotated open-ended questions. We introduce the Emergency Room Visual Question Answering (ERVQA) dataset, consisting of <image, question, answer> triplets covering diverse emergency room scenarios, a seminal benchmark for LVLMs. By developing a detailed error taxonomy and analyzing answer trends, we reveal the nuanced nature of the task. We benchmark state-of-the-art open-source and closed LVLMs using traditional and adapted VQA metrics: Entailment Score and CLIPScore Confidence. Analyzing errors across models, we infer trends based on properties like decoder type, model size, and in-context examples. Our findings suggest the ERVQA dataset presents a highly complex task, highlighting the need for specialized, domain-specific solutions.