Multimodal ChatGPT for Medical Applications: an Experimental Study of GPT-4V

TL;DR

This study evaluates GPT-4V's performance on medical visual question answering across diverse datasets, revealing its current limitations and unreliability for real-world diagnostic applications.

Contribution

It provides a comprehensive assessment of GPT-4V's capabilities in medical VQA, highlighting its constraints and informing future improvements.

Findings

GPT-4V shows unreliable accuracy in medical diagnostics.

The model's behavior has seven distinct limitations.

GPT-4V is not recommended for real-world medical diagnostics.

Abstract

In this paper, we critically evaluate the capabilities of the state-of-the-art multimodal large language model, i.e., GPT-4 with Vision (GPT-4V), on Visual Question Answering (VQA) task. Our experiments thoroughly assess GPT-4V's proficiency in answering questions paired with images using both pathology and radiology datasets from 11 modalities (e.g. Microscopy, Dermoscopy, X-ray, CT, etc.) and fifteen objects of interests (brain, liver, lung, etc.). Our datasets encompass a comprehensive range of medical inquiries, including sixteen distinct question types. Throughout our evaluations, we devised textual prompts for GPT-4V, directing it to synergize visual and textual information. The experiments with accuracy score conclude that the current version of GPT-4V is not recommended for real-world diagnostics due to its unreliable and suboptimal accuracy in responding to diagnostic medical questions. In addition, we delineate seven unique facets of GPT-4V's behavior in medical VQA, highlighting its constraints within this complex arena. The complete details of our evaluation cases are accessible at https://github.com/ZhilingYan/GPT4V-Medical-Report.