ModaLens: Measuring Image Sensitivity in Report-Conditioned Medical VLMs

2026-09-14

AI Summary

研究使用ModaLens方法通过图像交换审计来衡量医学视觉-语言模型在有无报告情况下的图像敏感性变化，揭示报告可用性降低了图像敏感性。

Abstract

A radiology report can already answer a clinical question, so it is hard to tell whether a vision-language model also uses the image. ModaLens, a paired image-swap audit, measures how report availability changes image sensitivity: MedGemma-27B on 3,199 paired MIMIC-CXR cases from 293 patients, all 14 questions per case (13 finding-specific and one composite), each image replaced by one from another study, usually of the same patient, with question and report fixed. Under an explicit answer instruction, the model's generated answer changes on 4.26 percent of trials with the report and 20.94 percent without it, a paired increase of 16.7 points (patient-clustered 95 percent CI 15.6 to 17.7), so report availability reduces image-swap sensitivity under this protocol; the original prompt with a lowercase first-token readout gives 4.70 percent against 17.07 percent, and substitutions also move continuous answer scores where the binary prediction does not change. The labels are derived from reports, which limits conclusions about visual correctness; the direction replicates in two further model lineages. Code, the exact prompts and a run record for every number are at https://github.com/criticaldata/MODALENS.

Problem

• vision-language model
• image sensitivity
• radiology report
• audit

Innovation

• ModaLens
• image sensitivity
• report-conditioned
• vision-language models
• medical VLMs