MedObvious: Exposing the Medical Moravec's Paradox in VLMs via Clinical Triage

TL;DR

MedObvious introduces a comprehensive benchmark to evaluate the critical pre-diagnostic sanity checks in medical vision-language models, revealing significant reliability issues in current models' ability to verify input consistency and safety.

Contribution

The paper presents MedObvious, a novel benchmark with 1,880 tasks for assessing input validation in medical VLMs, highlighting the need for safety-critical verification capabilities.

Findings

Models often hallucinate anomalies on normal inputs.

Performance drops with larger image sets.

Accuracy varies between multiple-choice and open-ended formats.

Abstract

Vision Language Models (VLMs) are increasingly used for tasks like medical report generation and visual question answering. However, fluent diagnostic text does not guarantee safe visual understanding. In clinical practice, interpretation begins with pre-diagnostic sanity checks: verifying that the input is valid to read (correct modality and anatomy, plausible viewpoint and orientation, and no obvious integrity violations). Existing benchmarks largely assume this step is solved, and therefore miss a critical failure mode: a model can produce plausible narratives even when the input is inconsistent or invalid. We introduce MedObvious, a 1,880-task benchmark that isolates input validation as a set-level consistency capability over small multi-panel image sets: the model must identify whether any panel violates expected coherence. MedObvious spans five progressive tiers, from basic orientation/modality mismatches to clinically motivated anatomy/viewpoint verification and triage-style cues, and includes five evaluation formats to test robustness across interfaces. Evaluating 17 different VLMs, we find that sanity checking remains unreliable: several models hallucinate anomalies on normal (negative-control) inputs, performance degrades when scaling to larger image sets, and measured accuracy varies substantially between multiple-choice and open-ended settings. These results show that pre-diagnostic verification remains unsolved for medical VLMs and should be treated as a distinct, safety-critical capability before deployment.