搜索医学 AI 论文与资源

ICLR 2026 Poster accepted paper at ICLR 2026. Generalist multimodal large language models (MLLMs) have achieved impressive performance across a wide range of vision-language tasks. However, their performance on medical tasks—particularly in zero-shot settings where generalization is critical—remains suboptimal. A key research gap is the limited understanding of why medical MLLMs underperform in medical image interpretation. **In this work**, we present a pioneering systematic investigation into the visual grounding capabilities of state-of-the-art medical MLLMs. To disentangle *visual grounding* from *semantic grounding*, we design VGMED, a novel evaluation dataset developed with expert clinical guidance, explicitly assessing the visual grounding capability of medical MLLMs. Code/project link: https://guimeng-leo-liu.github.io/Medical-MLLMs-Fail/

ICLR 2026 Poster accepted paper at ICLR 2026. Vision-Language Models (VLMs) for radiology report generation are typically trained to mimic the narrative flow of human experts. However, we identify a potential limitation in this conventional paradigm. We hypothesize that optimizing for narrative coherence encourages models to rely on linguistic priors and inter-sentence correlations, which can weaken their grounding in direct visual evidence and lead to factual inaccuracies. To investigate this, we design a controlled experiment demonstrating that as textual context increases, a model's reliance on the input image systematically decays. We propose LLaVA-TA (Topic-guided and Anatomy-aware), a new fine-tuning framework that directly addresses this challenge by re-engineering the generation process.

ICLR 2026 Poster accepted paper at ICLR 2026. Large visual language models (VLMs) have shown strong multi-modal medical reasoning ability, but most operate as end-to-end black boxes, diverging from clinicians’ evidence-based, staged workflows and hindering clinical accountability. Complementarily, expert visual grounding models can accurately localize regions of interest (ROIs), providing explicit, reliable evidence that improves both reasoning accuracy and trust. In this paper, we introduce **CARE**, advancing **C**linical **A**ccountability in multi-modal medical **R**easoning with an **E**vidence-grounded agentic framework. Unlike existing approaches that couple grounding and reasoning within a single generalist model, CARE decomposes the task into coordinated sub-modules to reduce shortcut learning and hallucination: a compact VLM proposes relevant medical entities; an expert entity-referring segmentation model produces pixel-level ROI evidence; and a grounded VLM reasons over the full image augmented by ROI hints.

ICLR 2026 Poster accepted paper at ICLR 2026. The growing demand for prenatal ultrasound imaging has intensified a global shortage of trained sonographers, creating barriers to essential fetal health monitoring. Deep learning has the potential to enhance sonographers' efficiency and support the training of new practitioners. Vision-Language Models (VLMs) are particularly promising for ultrasound interpretation, as they can jointly process images and text to perform multiple clinical tasks within a single framework. However, despite the expansion of VLMs, no standardized benchmark exists to evaluate their performance in fetal ultrasound imaging. Code/project link: https://github.com/BioMedIA-MBZUAI/FETAL-GAUGE

ICLR 2026 Poster accepted paper at ICLR 2026. Large language models perform well on many medical QA benchmarks, but real clinical reasoning is harder because diagnosis often requires integrating evidence across multiple images rather than interpreting a single view. We introduce MedThinkVQA, an expert-annotated benchmark for thinking with multiple images, in which models must interpret each image, combine cross-view evidence, and solve diagnostic questions under intermediate supervision and step-level evaluation. The dataset contains 10,067 cases, including 720 test cases, with an average of 6.68 images per case, substantially denser than prior work (earlier maxima $\leq$ 1.43). On the test set, the best closed-source models, Claude-4.6-opus, Gemini-3-pro, and GPT-5.2-xhigh, achieve only 54.9%--57.2% accuracy, while smaller proprietary variants, GPT-5-mini/nano, drop to 39.7% and 30.8%.

系统研究医学 MLLM 在医学图像视觉定位中的失效模式，提出 VGMED 评估数据集与 VGRefine 推理时方法，面向医学视觉问答与医学图像解释场景。