Medical Vision-Language Models

Diabetic Retinopathy (DR) is an aggressive retinal disease and a leading cause of global blindness, yet its clinical management is currently hindered by the black-box nature of diagnostic AI. While deep learning models achieve high classification accuracy, there is a critical lack of explainability methods capable of detailing the exact anatomical landmarks and lesion distributions that lead to a clinical decision for DR. Therefore, we propose HSQ-VLM, a novel quadrant segmentation pipeline on fundus images that utilizes a Landmark-Anchored Cartesian Cross-Attention mechanism to unify visual feature extraction with structured clinical reasoning. Unlike traditional methods that rely on arbitrary image partitioning, our pipeline implements 4-quadrant Topological Latent Partitioning (TLP) to dynamically align retinal features with a fovea-centered coordinate system. This allows the Vision-Language Model to generate natural language reports that quantify pathology with anatomical precision. On a dataset of 3,500 high-resolution fundus images, this innovative methodology achieved a lesion detection sensitivity of 99.6% for hemorrhages and 96.4% for microaneurysms, while demonstrating a significant reduction in boundary-ambiguity errors compared to standard segmentation baselines.

We present OpenMedQ, a medical vision-language model pretrained on the broadest fully-open medical mix to date: 14 datasets totaling ~3.35M pretraining samples spanning pathology, radiology, microscopy, and text-only clinical QA. OpenMedQ reaches state-of-the-art BLEU-1 on PathVQA (75.9), beating Med-PaLM M variants up to 562B parameters (~80x larger), and matches the best reported VQA-MED BLEU-1 (64.5). Its vision encoder, transferred to 8 unseen medical classification benchmarks under an identical downstream recipe, obtains the highest average macro-F1 (0.757) among BiomedCLIP (0.745), PMC-CLIP (0.745), PubMedCLIP (0.746), and a from-scratch baseline (0.616). We release our code and an interactive demo is publicly available as a reproducible baseline for the community.

High-stakes clinical use of large vision-language models (LVLMs) requires reasoning that is grounded in visual evidence and clinical knowledge, not just correct final answers. We introduce OpenMedReason, a large-scale, open multimodal medical reasoning corpus comprising approximately 450K image-question-answer instances whose reasoning traces are primarily derived from curated biomedical, human-authored scientific articles. OpenMedReason provides high-fidelity supervision beyond synthetic chains of thought, covering diverse medical domain vision modalities such as radiological scans, microscopic images, visible light photographs, charts, and others. We complement it with OpenMedReason-Bench, a held-out benchmark that allows fine-grained evaluation of LVLMs along three complementary axes of capability, including perception, medical knowledge, and rationale, enabling diagnostic evaluation beyond final-answer accuracy. OpenMedReason is a rich training resource that exhibits its effectiveness in both supervised fine-tuning (SFT) and reinforcement-based alignment. Training with OpenMedReason yields a 20% average improvement in VQA accuracy over the base model and achieves performance within 4.2% of the strongest comparable-scale medical LVLMs. Fine-grained performance analysis confirms that the gains are not concentrated in any single axis: OpenMedReason improves perception, medical knowledge, and rationale jointly, and its reasoning traces are preferred over those of the base model in 86.1% of pairwise comparisons. We release the code and dataset at huggingface.co/datasets/neginb/OpenMedReason.

3D vision-language segmentation aims to segment target objects in 3D scenarios according to the linguistic instructions and visual observations. Prior art heavily relies on the coarse superpoint representation to reduce the computation complexity, which suffers from poor segmentation quality and messy object boundaries. In this paper, we propose the SEGment-And-select (SEGA3D) paradigm for 3D visionlanguage segmentation that directly operates on the fine-grained visual information and is free from the superpoint dependency. Specifically, we first leverage a mask candidate generator to provide fine-grained categorical mask candidates, substantially improving the quality of candidate masks over the superpoint counterparts. Then, a Large Language Model (LLM) is utilized to generate the semantic and spatial information based on the linguistic description and visual features. The LLM output and visual features are fed to the Semantic-Spatial Selector (SSS) to produce the top-ranking mask candidates. Eventually, the Loopback Verification Module (LVM) is designed to yield the segmentation mask from the selected candidate masks. Our SEGA3D attains competitive performance on ScanRefer, ScanNet and Matterport3D benchmarks. Notably, our SEGA3D surpasses the top-performing counterpart by 8.3 mIoU and 5.3 mIoU on ScanNet and Matterport3D, respectively. Codes will be available upon publication.

Medical vision-language models (VLMs) are evaluated on public benchmarks whose images and question-answer pairs have been freely downloadable for years, yet reported accuracy assumes these examples were absent from pretraining. We audit open VLMs on SLAKE-En, PathVQA, VQA-RAD, and an auxiliary public OmniMedVQA mirror using four detector families: image-side near-neighbour overlap against PMC-OA-beta, canonical-order exchangeability, cohort-relative Min-K%++ tail enrichment, and cross-model top-K overlap. We find measurable image-side source overlap on SLAKE-En: 19.8% of images are flagged under SigLIP-B-16 and 4.2% under SigLIP-SO400M, while out-of-domain controls produce 0/2000 flags. Manual adjudication shows same-modality, same-projection matches to different patients rather than verified pixel-level duplicates, so we interpret this as source or distributional overlap rather than confirmed per-image memorization. On the text side, Qwen2.5-VL on SLAKE-En shows a canonical-order exchangeability signal that survives ordering ablation and external non-medical baselines. On the OmniMedVQA mirror, exchangeability fires for five medical and general VLMs while BLIP-2 remains clean. In contrast, cohort-relative Min-K%++ tail enrichment and cross-model top-K overlap collapse under an external pre-domain baseline: BLIP-2 reproduces the apparent positive signals despite lacking plausible medical-VQA exposure. We conclude that these cohort-relative detectors are unreliable as standalone membership-inference signals on small medical-VLM cohorts.

Vision-language models (VLMs) pretrained on large-scale image-text pairs demonstrate strong image-level understanding, but are primarily optimized for global alignment and do not explicitly encode fine-grained anatomical structure, limiting their suitability for spatially precise tasks such as segmentation. We introduce CheXanatomy, a framework that integrates explicit anatomical knowledge into a pretrained VLM through autoregressive token-space supervision. Instead of adding task-specific decoder heads, the model is trained to generate anatomical segmentation masks via next-token prediction. To enable scalable supervision, we synthesize realistic chest radiographs from CT volumes and forward-project CT segmentation labels to obtain anatomically consistent 2D masks. We evaluate the approach on synthetic and real chest radiographs against a U-Net baseline, including ablations on model scale, input resolution, and vision encoder fine-tuning. Autoregressive anatomical supervision achieves performance comparable to specialized convolutional models in-distribution and demonstrates improved geometric robustness under domain shift to real CXR data. In addition, anatomy-pretrained models exhibit improved sample efficiency when adapting to novel localization tasks under limited supervision. Larger models and higher input image resolution improve performance, while vision encoder fine-tuning has limited effect. These results show that embedding anatomical structure directly into the generative objective promotes spatially grounded representations and supports anatomy-aware medical vision-language modeling.

Objectives: Automatic data extraction from free-text radiology reports enables large-scale research, but few studies assessed the performance of large language models (LLMs) on Dutch neuroradiology reports. Methods: We analyzed 947 brain MRI reports from a tertiary memory clinic (2016-2021), authored by consultant neuroradiologists. Trained medical students annotated thirty variables; 100 reports were double-annotated to assess inter-rater reliability. We evaluated the performance of the open-weight LLM LLaMA 3.1 using different languages (Dutch vs. English translation) and few-shot prompting with different example selection strategies. Performance was evaluated using balanced accuracy for categorical variables, accuracy and mean absolute error for counts, and text similarity for free-text. Metrics were computed across 10 random splits of the 947 reports. Results: LLaMA 3.1 demonstrated high zero-shot performance for visual rating scores (mean [95%-CI]): Medial Temporal Atrophy: 90% [77-100%] on the left and 96% [94-99%] on the right, Global Cortical Atrophy: 87% [83-91%], and Fazekas: 94% [93-96%]. Microbleed mentions were detected with 93% accuracy [92-95%] and infarct mentions with 82% [80-84%]. Text similarity for lesion location reached 0.95 [0.95-0.96]. Performance was lower for numerical variables: 80% [78-82%] for the number of microbleeds and 66% [63-68%] for infarcts. English translation yielded comparable results. Few-shot prompting improved performance for numerical variables, achieving 92% [90-93%] for microbleeds and 81% [77-85%] for infarcts using structural similarity-based selection. Conclusion: LLaMA 3.1 shows strong potential for extracting data from Dutch neuroradiology reports. Few-shot prompting enhances performance for numerical variables, whereas challenges remain for location-specific variables.

Despite advances in 3D scene understanding, existing 3D Large Multimodal Models operate in offline settings, requiring complete scene observations or predefined video clips. In this paper, we present an online 3D vision-language model that enables real-time spatial understanding from streaming video. Our approach adopts an autoregressive streaming control modeling based on the LLM's next-token prediction objective to learn when to respond, and employs a lightweight Visual-Spatial Feature Integration (VSFI) module to incrementally inject temporally aligned geometry priors into the visual stream. To alleviate long-context decoding overhead, we propose a plug-and-play Geometry-Adaptive Voxel Compression (GAVC) module for efficient visual token compression. To address the scarcity of streaming 3D-language data, we further develop a scalable data generation pipeline that curates over 1M online spatio-temporal 3D QA pairs and establishes a comprehensive benchmark spanning 29 tasks. Extensive experiments show that our approach significantly outperforms both proprietary and open-source models across online and offline 3D spatial understanding, reasoning, and grounding tasks. The project page is available at https://stream3d-vlm.github.io/

Medical large vision-language models (Med-LVLMs) have recently achieved remarkable progress in vision-language comprehension and medical image segmentation. However, existing models still struggle to unify these two capabilities, which is essential for achieving clinically reasoning that connects visual findings with semantic interpretation. We present MedSIGHT, a unified framework that equips Med-LVLMs with structured, pixel-level understanding for grounded visual comprehension. MedSIGHT introduces a novel Region Perceiver module that produces region-centric tokens, encoding spatial information directly into representation space of the language model. We further propose a medical region codebook into the LLM vocabulary, allowing the model to generate discrete region codes as symbolic representations of anatomical and pathological regions. These codes are decoded through the Region Perceiver to reconstruct segmentation mask, achieving end-to-end spatial grounding. Lastly, MedSIGHT combines Region Perceiver, Codebook and LLM using our proposed progressive training strategy to gradually aligns these modules stably. Trained on only 72K multimodal instruction pairs, MedSIGHT achieves state-of-the-art performance across diverse imaging modalities on both medical comprehension and segmentation tasks.

Vision and language models (VLMs) hold immense promise to transform biomedical imaging workflows, from detecting lesions in chest X-rays to profiling cellular features in microscopy. Realizing this potential, however, requires robust and fine-grained visual perception. Models need to correctly interpret subtle features in images, and they must do so across diverse biomedical modalities, scales, and contexts. Nevertheless, current benchmarks remain limited. To address these gaps, we introduce the Massive Multimodal Biomedical Understanding (MMBU) benchmark. It is the largest biomedical vision and language benchmark to date, covering 35 submodalities with rich structured metadata. It includes both open and closed versions of ungrounded classification, grounded classification, and object detection, enabling systematic evaluation of model performance across biological scales, clinical settings, and imaging modalities. Evaluating 15 open-weight and 2 frontier VLMs, we find that while medical adaptation provides measurable gains for some models, the high accuracy often reported on established benchmarks can mask deficiencies in visual perception and domain generalization.

Medical vision-language models (VLMs) have shown increasing potential for clinical image interpretation, including lesion detection and report generation. However, their practical utility remains limited by insufficient sensitivity to subtle lesions, whose visual evidence is often sparse, low-contrast, and embedded within complex anatomical context. As local visual tokens are aggregated, these weak lesion cues can become underrepresented in global image representations, making them difficult for medical VLMs to recognize. Existing efforts to improve lesion sensitivity mainly rely on medical-domain vision-encoder pre-training, clinical-term-guided alignment, or trainable pathological representation enhancement. Although effective, these approaches usually require additional training or model-specific adaptation and may overfit to particular disease morphologies, limiting their applicability to frozen medical VLMs. To address these limitations, we propose EasyLens, a training-free plug-and-play subtle-lesion representation amplifier for medical VLMs. EasyLens first constructs EasyBank, a pathology-anatomy prototype space that provides lesion-related prototypes and anatomy-aware normal references for comparing suspicious patches against both pathological and normal anatomical patterns. To avoid blindly amplifying normal tissues, EasyTag selects lesion-relevant patches through counterfactual prototype reasoning. To counteract the dilution of subtle lesion cues in global image representations, EasyAmplifier strengthens the selected lesion-relevant patch representations through morphology-guided residual enhancement, thereby increasing their contribution to the global image embedding. Experiments on multiple medical image datasets and frozen medical VLM backbones show that EasyLens improves subtle-lesion detection and outperforms existing encoder-enhancement baselines.

Vision-Language Models (VLMs) struggle when applied to medical image-text data, yet the tools available to diagnose this failure remain limited. Existing representation alignment metrics are symmetric, collapsing both modalities into a single score and hiding which modality drives cross-modal degradation. We introduce the Spectral Alignment Score (SAS), an asymmetric metric that projects both modalities onto the principal eigenbasis of an anchor modality and computes eigenvalue-weighted per-eigenmode correlations, resulting in directional scores whose difference quantifies modality information imbalance. We embed SAS within a benchmarking framework evaluating 15 VLMs across natural and medical image-text datasets alongside 6 alignment metrics and bidirectional retrieval. Our experiments show that medical images retain richer structural information than their paired clinical reports, a directional asymmetry invisible to all competing metrics, and that SAS achieves the strongest zero-label correlation with retrieval performance in the medical domain, positioning it as a practical diagnostic tool for clinical deployment. Code is available at this URL: https://github.com/iamalegambetti/medical-vlms-assessment.

Medical Vision-Language Models (VLMs) are typically evaluated on English radiology visual question answering benchmarks, leaving their robustness under non-English clinical language largely unexplored. We introduce IndoRad-VQA, an Indonesian adaptation of VQA-RAD, to assess whether medical VLMs retain radiology reasoning ability when questions are asked in Bahasa Indonesia. Radiology question-answer pairs are translated into Indonesian with self-evaluation-based quality control to preserve clinical meaning, terminology consistency, and answer equivalence. We evaluate general-purpose, Southeast Asian multilingual, and medical-specific VLMs under English and Indonesian prompting settings. Beyond accuracy, we quantify the language robustness gap between English and Indonesian inputs. We also conduct an error analysis to identify failure modes of question answering, such as yes/no flips, laterality errors, and output-language mismatches. Our findings show that strong performance on English medical VQA benchmarks does not necessarily translate to robust behavior in Indonesian clinical contexts. We observe a performance gap of 8 to 25 percent between the English and Indonesian settings, depending on the evaluation metric. These results highlight the need for more inclusive multilingual evaluation of medical multimodal foundation models. The dataset is available at https://huggingface.co/datasets/Lab-IS/IndoRad-VQA.

Background: Oral diseases affect nearly 3.5 billion people worldwide, yet the comparative clinical potential of large-scale AI models in dentistry remains poorly understood. Three distinct model categories have emerged: language-generative models, discriminative vision foundation models, and dental-specific foundation models, with no unified review examining their relationships and collective limitations. Methods: Following PRISMA-ScR guidelines, we systematically searched four databases (PubMed, Google Scholar, Scopus, arXiv), screened independently by two reviewers. After applying inclusion/exclusion criteria, 97 studies (2020-2026) were included. We propose a two-dimensional classification framework organizing models by architectural paradigm and dental specialization degree. Results: Language-generative models excel at text-based tasks (clinical reasoning, licensing exams, patient communication) but show inconsistent performance on image-dependent diagnostics. Adapted SAM and CLIP variants achieve strong tooth segmentation and lesion detection results. Dental-specific models (DentVFM, DentVLM, OralGPT) demonstrate strongest performance on complex multimodal tasks. Integrated pipelines consistently outperform single-model approaches. A data asymmetry is observed: dental-specific pretraining concentrates almost entirely in the vision domain, reflecting scarce large-scale dental text corpora. Conclusions: General-purpose and dental-specific models play complementary roles; the most effective systems combine both within structured pipelines. Safe autonomous deployment requires resolving three persistent barriers: hallucination in generative models, limited annotated dental datasets, and absent standardized clinical evaluation benchmarks.

Vision-language models (VLMs) trained on paired chest radiographs and radiology reports learn a shared embedding space that can preserve instance-level image-report correspondence. This poses a privacy risk in settings where radiographs and reports are deliberately kept separate after acquisition, such as image-only data sharing or access-controlled reports, because a de-identified image may be re-linked to its original narrative report through cosine similarity alone. We formalized this as image-to-report retrieval and used public paired cohorts, in which the true pairing is known by design, as ground-truth benchmarks to audit the risk rather than as the privacy scenario. Evaluating VLMs of increasing clinical specialization on 406,241 paired examples from 126,804 patients across MIMIC-CXR (43,793 held-out pairs) and external CheXpert Plus (29,296 pairs), we found that re-linkage rose systematically with specialization: the strongest VLM retrieved the correct report at 15 times chance at a candidate pool of N = 100, 50 times chance at N = 10,000, and well above chance at full-database scale. The signal persisted under pathology-matched hard negatives that removed disease-label shortcuts, indicating correspondence beyond broad diagnostic categories. To reduce it without retraining, we froze both encoders and applied differentially private optimization only to the projection heads defining the alignment layer (epsilon = 0.34, delta = 6x10-6). This reduced Recall@1 by 61.8% at N = 10,000 on MIMIC-CXR and transferred to CheXpert Plus without retraining, while image-side utility was largely preserved: macro AUROC for linear-probe classification across 14 labels shifted only from 79.63% to 79.43%. Targeted DP finetuning of the shared alignment layer can substantially reduce cross-modal re-linkage without materially degrading the image representations that make these models clinically useful.

Medical imaging interpretation is a foundational pillar of modern clinical diagnostics, yet the manual generation of radiology reports remains a time-consuming process prone to interpretation inconsistencies. Within the field of medical AI, automating these descriptions through deep learning promises to streamline clinical workflows and standardise diagnostic output. However, accurate disease detection and precise report generation remain significant challenges due to limitations in capturing fine-grained visual features and ensuring clinical coherence. To address these issues, we propose RL-ACRGNet, an improved encoder-decoder model that integrates a pre-trained DenseNet encoder with a multilevel LSTM decoder within an off-policy reinforcement learning framework. Using a dual-network approach to refine visual-semantic embeddings through a metric-based reward mechanism, we demonstrate that RL-ACRGNet consistently outperforms state-of-the-art baselines on the IU-Xray dataset, achieving quantitative improvements in BLEU-4 (0.47%), METEOR (0.17%) and ROUGE-L (0.518). Furthermore, comprehensive evaluations on the large-scale MIMIC-CXR data set confirm the robust generalisation of the model and its ability to generate high-quality, clinically relevant reports

Learning transferable and interpretable representations from medical volumetric scans remains challenging due to complex anatomical structures and weak, heterogeneous supervision provided by radiology reports. In this paper, we propose Anatomy-aware Semantically-Adaptive Pre-training (ASAP), a principled vision-language pre-training framework for fine-grained medical volumetric representation learning from large-scale chest CT scans and their corresponding radiology reports. ASAP integrates three key components: (1) an anatomy-aware knowledge injection module that incorporates organ-level structural priors via off-the-shelf segmentation tool to encourage anatomically coherent representations; (2) a semantically-adaptive selective alignment mechanism that dynamically associates sentence-level findings with localized volumetric regions; and (3) a semantically-adaptive fusion module for effective interaction between anatomically informed visual features and grounded textual cues under dual-modal masked modeling paradigm. Beyond methodological contributions, we establish a comprehensive benchmark for medical volumetric vision-language pre-training on chest CT, covering 15 datasets and 22 downstream tasks spanning abnormality classification, segmentation, disease prognosis prediction, report generation, vocabulary classification, cross-modal retrieval and visual question answering. This benchmark provides standardized evaluation protocols to systematically assess representation quality under diverse clinical settings and data regimes. Extensive experiments demonstrate that ASAP consistently achieves state-of-the-art performance across tasks and datasets, with particularly pronounced gains under limited supervision and distribution shift, validating its effectiveness in learning transferable and clinically meaningful volumetric representations.

Reinforcement learning with verifiable rewards has rapidly advanced reasoning in vision--language models. However, for chest X-ray report generation, the standard rewards (i.e. exact-match accuracy and step-level processes) are incompatible because the reports consist of unordered and orthogonal findings, rather than a causal reasoning chain. We address this gap with a set-based view: each report is split into sentences and embedded by a frozen sentence transformer, yielding unordered embedding sets. We propose the use of set-to-set distances between generated and reference embeddings as continuous, permutation-invariant rewards. Across two datasets and three vision--language models (Qwen3-VL-2B/4B, Gemma3-4B), post-training with set-to-set distance based rewards via GRPO consistently outperforms supervised fine-tuning and exact-match GRPO on all headline metrics (BERTScore, RadGraph F1 and CheXbert F1 by average %6.80, %7.82 and %4.45 relative improvements respectively). The same set distances also enable test-time best-of-$N$ selection: scoring candidates by their distance to training-report embeddings outperforms random selection on our trained models as well as three closed-source LLMs (Mistral-Small, Gemini-2.5 Flash-Lite, GPT-4o-mini) with on average %16.4 relative improvement on BERTScore. Used as a streaming signal, they support a more efficient form of test-time scaling: pruning low-scoring candidates mid-generation reduces generated tokens by over 50% while preserving the Findings quality of full best-of-$N$ selection. Together these results establish set-distance rewards as a unified signal for both post-training and test-time scaling in chest X-ray report generation. Our code is publicly \href{https://anonymous.4open.science/r/Set-Distance-Rewards-CXR-BFDA}{available}.

Modern 3D medical vision-language models (VLMs) can generate fluent radiology-style text while exhibit critically low pathology detection and output diversity, collapsing to generic templates that under-report rare yet critical findings. We identify this failure mode as Template Collapse. This failure stems from the unique constraints of 3D medical imaging, e.g., limited data, severe label imbalance, and weak signals from volumetric encoders. Under these constraints, text-generation objectives encourage shortcut learning and fluent but weakly grounded reports. We systematically diagnose the Template Collapse through clinical fidelity, output diversity, normal-template bias, and rare-finding survival. To mitigate it, we propose CLarGen, a decoupled framework that separates what to say (clinical detection) from how to say it (language synthesis). CLarGen uses (i) a Latent Query Transformer for multi-label pathology detection, (ii) pathology-guided retrieval for clinically matched exemplars, and (iii) a medical language model to synthesize the final report from detected findings and retrieved context. Across state-of-the-art 3D CT report generation baselines, CLarGen mitigates Template Collapse and substantially improves clinical accuracy (macro-F1 0.487 vs. 0.189; CRG 0.472 vs. 0.368) while maintaining fluent reporting. Our results suggest that explicit, measurable clinical grounding is essential for template-collapse-resistant 3D CT report generation. Code is available at https://github.com/ai-med/CLarGen.

Generating clinically useful pathology reports for pathology cases from whole-slide images (WSIs) is challenging due to gigapixel resolution, long visual-token sequences, and the complexity of case-level reasoning, where a single case may contain multiple WSIs with heterogeneous tissues and ambiguous findings. We present a simple token-efficient vision--language model for case-level synoptic report generation that remains practical under constrained GPU memory. Our architecture follows a minimal three-component design: a frozen pathology patch encoder, a lightweight two-layer MLP vision-language aligner, and a large language model decoder, with an explicit WSI marker token to separate slides within a case. Training proceeds in two supervised stages: (1) aligner-only WSI captioning using heterogeneous WSI-text pairs, and (2) case-level supervised fine-tuning on case-report pairs for structured report generation. To reduce sequence length, we represent each slide using $512 \times 512$ patches at $5\times$ magnification, which reduces the average sequence length by up to $64\times$ times compared to the commonly used $20\times$ patches. Combined with efficient training techniques, we enable practical training with only half a NVIDIA H100 GPU. Across both training stages, our approach achieves high ROUGE-L/METEOR/BLEU-4 scores while being substantially more efficient in memory and runtime. In AI-based evaluations, our model is consistently preferred over strong baselines. Extensive ablations characterize performance-efficiency trade-offs and identify simple choices that improve robustness in multi-WSI settings. Overall, this work provides a strong, reproducible baseline for efficient pathology report generation, lowering the barrier to multi-WSI VLM research under limited compute.

We present GR3D, a spatial vision language model equipped with three complementary grounding capabilities--explicit 2D grounding, implicit 2D grounding, and monocular 3D grounding--within a single framework. GR3D introduces an implicit grounding mechanism that identifies entity mentions during generation and inserts the corresponding region tokens into the text stream, allowing the model to reference visual evidence on the fly when producing spatial chain-of-thought responses. In parallel, a region-prompted monocular 3D grounding design predicts 3D bounding boxes in the camera view from grounded region queries, supported by intrinsic-aware normalization and dense geometric supervision. Together, these grounding capabilities enable GR3D to decompose complex spatial understanding problems into grounded 2D perception followed by 3D inference. GR3D achieves consistent improvements across grounded and non-grounded spatial benchmarks, demonstrating grounding as an effective inductive bias for strengthening spatial understanding in VLMs. These grounding capabilities collectively enhance general spatial understanding beyond the grounding task itself.

Multimodal Large Language Models (MLLMs) have shown strong performance on public medical benchmarks, yet existing evaluations often remain weak proxies for clinical use, relying on isolated inputs and simplified recognition-style tasks. We introduce CardioLens, a leakage-resistant evaluation testbed for multi-sequence Cardiovascular Magnetic Resonance (CMR), constructed from private hospital archives through a rigorous report-to-QA construction and verification pipeline. CardioLens contains 473,896 slices and 13,494 verified QA pairs across 4D Cine, LGE, perfusion, and T2-weighted imaging, and evaluates three stages of CMR interpretation: image understanding, report generation, and disease diagnosis. Across 24 state-of-the-art MLLMs, CardioLens reveals a substantial clinical reality gap: models perform poorly overall, with performance degrading along the real CMR workflow. Confusion analysis further shows a category-collapse failure mode, where models default to frequent abnormal categories rather than distinguishing clinically distinct findings. To rule out MLLM-compatible input construction as the primary cause, we compare random, clinically motivated, and data-driven slice selection protocols under different slice budgets; performance changes only marginally, typically by about 1%. Explicit reasoning prompts also fail to rescue performance, often making models more conservative rather than improving visual evidence use. These results show that current MLLMs remain far from reliable CMR interpretation, where clinical decisions require integrating distributed evidence across sequences, views, and temporal phases. CardioLens provides a clinically grounded testbed for developing next-generation MLLMs toward real-world clinical deployment.

Evaluations of dental vision-language models remain fragmented across datasets, task definitions and metrics, and often ignore their computational cost. This limits their widespread deployment for dental screening outside specialist centres, where timely inference, limited hardware, and local handling of patient images are vital for practical, privacy-preserving clinical prescreening. Here we present Pocket-Dentist, an efficiency-aware benchmark for dental multimodal question answering that brings together three datasets spanning approximately 1,159 patients from BRAR and MetaDent, five task types and seven metrics. Across 14 typical VLMs, our results reveal an interesting observation: compact VLMs, such as 2B-parameter models, become competitive with much larger VLMs on most metrics after lightweight adaptation while requiring substantially lower computational costs in dental image understanding. Deployed locally on an iPhone 17 Pro, our finetuned compact VLM Pocket-Dentist-2B processed each sample in 4.31 s, reducing latency by 4.9x and memory use by 2.3x compared with a 7B baseline. Our project page is available at https://2026-icml.github.io/pocket-dentist-icml.

Volumetric Reasoning Segmentation (VRS) aims to segment a target region in a 3D medical scan from a free-form clinical query, where the referent is often implicit and requires both medical knowledge and volume-grounded reasoning. Existing methods typically rely on specialized segmentation tokens to connect language with mask decoding, but this coupling collapses the decision process into opaque latent representations, limiting interpretability and generalization to diverse narrative expressions. In this paper, we present MedVol-R1, a reinforcement learning-based framework for VRS that explicitly decouples evidence grounding from volumetric delineation: the LVLM grounds clinical reasoning to a verifiable 2D evidence anchor (key axial slice and 2D bounding boxes), which is then propagated into a coherent 3D mask by a frozen MedSAM2 module. We train MedVol-R1 with cold-start supervised fine-tuning followed by GRPO, guided by a multi-component reward that encourages informative evidence selection, accurate 2D spatial grounding, and cross-slice volumetric coherence, without requiring costly chain-of-thought annotations. Experiments on CT-ORG, AbdomenCT-1K, and KiTS23 from the M3D-Seg benchmark demonstrate that MedVol-R1 consistently outperforms strong baselines and achieves state-of-the-art performance, with reinforcement learning providing clear gains over pure supervised fine-tuning.

Modern deep learning offers powerful tools for automated retinal screening, but it remains unclear how different visual model families compare in realistic multi-disease settings and under domain shift. In this work, we benchmark twelve architectures across four model families: convolutional neural networks, vision transformers, hybrid CNN-transformer backbones, and vision-language models, using the Retinal Fundus Multi-disease Image Dataset (RFMiD). We evaluate two tasks: binary screening for any retinal disease and multi-label classification across 28 disease classes. Using standardized training, calibration, and evaluation protocols, we report AUC, F1, precision, recall, and sensitivity at a clinically relevant operating point with specificity near 80%. On RFMiD, all architectures perform well on binary screening, with AUC above 84%, but attention-based models perform best. SwinTiny and the hybrid CoAtNet0 and MaxViTTiny models achieve the strongest binary screening results and improve macro and micro F1 in the multi-label setting. Vision-language models, including CLIP ViT-B/16 and SigLIP-Base384, are competitive with CNN baselines but do not surpass the best transformer and hybrid backbones. In external validation on Messidor-2 for referable diabetic retinopathy, AUC ranges from 66.8% to 84.7%, with hybrid and transformer models again showing strong performance. These results provide a reproducible reference for model selection in multi-disease retinal screening and guide future automated screening tools for clinical deployment.

Urgent suspected colorectal cancer (CRC) referrals create operational bottlenecks because semi-structured clinical documents often require manual review and transcription. The original RAPTOR system used Large Language Models for structured extraction but relied on a separate OCR stage, making it vulnerable to handwriting, layout variation, and loss of visual evidence linkage. We present RAPTOR+, a multimodal extension that uses Vision-Language Models (VLMs) for end-to-end referral understanding. We evaluate fine-tuned VLMs, commercial and open-source zero-shot VLMs, and the original OCR-based pipeline on 223 clinically curated CRC urgent referral forms. We also introduce a grounding-aware evaluation framework that measures both extraction accuracy and evidence localisation. Results show a clear grounding gap in zero-shot models. Gemini 2.5 Flash achieved 92.6% Reading Accuracy but only 1.2% Strict Safety. In contrast, fine-tuned Qwen3-VL-8B achieved 96.1% Reading Accuracy and 60.6% Strict Safety, substantially improving verifiable evidence grounding. These findings show that task-specific fine-tuning is essential for reliable, auditable clinical document understanding. RAPTOR+ enables extracted referral decisions to be linked to visual evidence, supporting safer and more efficient cancer referral triage.

Medical vision-language models (VLMs) often hallucinate findings when generating chest X-ray reports: they fabricate findings that are not present in the image, miss important ones, or locate them incorrectly. We mitigate this without weight updates by decoding-time residual steering on a per-token sparse autoencoder (SAE) basis: Top-$K$ SAEs on late layers, causal steering against clinical errors, then combined suppress/boost intervention at inference time. On the MIMIC-CXR test split, our inference-only method improves the quality of generated reports for three radiology VLMs (RadVLM, LLaVA-Rad, and CheXOne), with relative improvements of +5.4%, +7.2%, and +17.0% in the clinical composite metric, and statistically significant GREEN gains on all backbones. A cross-model feature alignment shows that the quality-promoting (boost) directions overlap strongly across architectures, whereas hallucination-linked (suppress) directions are model-specific. Therefore, transferable steering must treat suppression per-backbone, rather than sharing a universal suppress list. The same recipe transfers zero-shot to IU-Xray (Green $+7.7\%$ rel.) without retraining, confirming that the identified features are properties of the model, not of the training corpus. We release causal feature sets and an interactive feature dashboard: https://cxr-sparse-feature-dashboard.netlify.app/.

Pre-trained Vision Transformers (ViTs) are increasingly deployed for medical image classification. However, correcting their inevitable failure cases in dynamic clinical scenarios poses a critical challenge. Conventional fine-tuning approaches inherently suffer from catastrophic forgetting, severely degrading previously acquired diagnostic capabilities. Such instability fundamentally compromises clinical safety. Addressing this vulnerability requires an active, controllable, and reliable intervention mechanism that is both theoretically grounded and inherently interpretable. To this end, we propose X-Edit (eXact, eXplicit, and eXplainable Editing), an efficient null-space model editing framework. X-Edit transitions the editing process from iterative gradient-based optimization to a theoretically grounded, closed-form solution. Specifically, we first explicitly localize the influential layers via causal tracing governing the erroneous prediction. Subsequently, we construct an orthogonal null-space projection matrix from a curated anchor set. By geometrically constraining the exact parameter update strictly within this null space, we provide mathematical guarantees that the intervention rectifies targeted errors without perturbing established diagnostic representations. Extensive evaluations on six medical imaging benchmarks demonstrate that X-Edit comprehensively suppresses catastrophic forgetting while achieving superior edit success rates. Our code is available at https://github.com/HenryLau7/X-Edit.

Classical noisy-label theory predicts that downstream performance under weak supervision is bounded above by the labeler's accuracy, implying a sharp crossover: once a gold-trained classifier matches the labeler, weak labels stop helping and start hurting. The prediction is theoretical; what is missing is a benchmark calibration that turns it into an instance-level statement for modern foundation-model labelers. We provide such a calibration for BiomedCLIP-generated weak labels on three medical-imaging benchmarks (PCAM, ISIC, NIH-CXR) and six downstream architectures spanning an 11x parameter range. The crossover predicted by theory appears at ng~100 on PCAM, 20-50 on ISIC, and 250-500 on NIH-CXR; weak labels above the crossover degrade AUC by up to -0.10. The location is architecture-invariant for four of five pretrained architectures, and a within-family DenseNet sweep (2.5x parameters, identical pretraining) supports the view that the labeler, not the student, is the dominant constraint. The calibration in turn produces a decision rule operable from 10-20 gold labels: compare gold-only AUC to VLM accuracy on the user's gold set. A structured-vs-random noise sign flip on NIH-CXR shows that the rate-only formulation of the bound is incomplete and identifies a concrete refinement (label-space projection) that future benchmarks can be designed to test.

CT report generation (CTRG) requires models to summarize three-dimensional anatomical context and pathological findings from hundreds of axial slices. Existing methods typically learn a direct image-to-text mapping, providing limited mechanisms for modeling how CT evidence evolves across slices or how reports respond to controlled changes in latent lesion-related factors. We propose SliceWorld, a CT-specific world-state framework that treats an axial CT scan as an ordered sequence along the z-axis. SliceWorld encodes prefix CT evidence into factor-aware latent states containing anatomy, lesion, and uncertainty components, and projects these states into world tokens used for multi-step future-slice feature prediction, lesion-factor intervention, and LLM-based report generation. The model is first pretrained on CT slice sequences with predictive, factor-aware, and counterfactual objectives, and is then fine-tuned on paired CT-report data. Experiments on M3D-Cap and CT-RATE show that SliceWorld improves natural language generation metrics and clinically oriented automatic evaluation. Further analyses demonstrate multi-horizon future-slice prediction, measurable factor alignment, reduced-slice robustness, and selective lesion-sensitive report modulation.

Grounding radiology report descriptions to 3D CT volumes is essential for verifiable clinical interpretation, yet remains challenging due to the semantic-spatial gap between free-text narratives and volumetric anatomy. Existing report-assisted and vision-language grounding methods typically rely on phrase-level alignment or dense pixel supervision, resulting in limited lesion-wise correspondence and suboptimal localization accuracy. We propose GLeVE, a graph-guided lesion grounding framework with anatomical prior verification and octree-based autoregressive refinement. GLeVE treats each lesion description as an atomic semantic unit and encodes organ attribution, attributes, and inter-lesion relations through relation-aware graph reasoning to produce discriminative lesion-wise queries. Anatomy-aware proposal generation with region-level verification enforces one-to-one text-lesion alignment, while hierarchical octree refinement progressively improves boundary delineation. Experiments on AbdomenAtlas 3.0 demonstrate consistent gains over classical multimodal foundation models and report-supervised baselines in both segmentation accuracy and lesion-level localization.

Vision-language models (VLMs) are increasingly used for medical image interpretation, yet they frequently hallucinate, generating clinically plausible but factually incorrect findings that pose direct patient safety risks. We introduce HalluCXR, a benchmark evaluating six architecturally diverse VLMs across 856 stratified MIMIC-CXR chest radiographs and three query types, yielding 15,408 model evaluations. An eight-category hallucination taxonomy with clinical severity ratings and a two-layer detection pipeline are validated against 250 human annotations (auto-detection F1=0.959; LLM judge F1=0.907). We find that 61.9--82.3% of outputs contain hallucinations, with clinically dangerous errors in up to 80.2%. Three key patterns emerge: normal radiographs paradoxically attract the most severe hallucinations, common findings are systematically over-fabricated while rare findings go under-detected, and response length alone predicts hallucination risk (AUC up to 0.908). A six-model ensemble reduces fabrication by up to 84.8% at the cost of increased omission; a three-model subset retains comparable performance at half the cost. These results establish that hallucination auditing, verbosity-based risk monitoring, and ensemble-based safety layers are prerequisites for clinical deployment.

Large Vision Language Models (LVLMs) show promise in medical applications, but their inability to faithfully ground responses in visual evidence raises serious concerns about clinical trustworthiness. While visual attribution methods are widely used to explain LVLM predictions, whether these explanations actually reflect the visual evidence underlying the model's decision is largely unverified, since ground-truth annotations for internal model reasoning are typically unavailable. We address this question for chest X-ray (CXR) reasoning by developing a causal evaluation framework that retains only CXR-VQA samples for which the expert-annotated region is verified, via counterfactual editing, to be causally responsible for the model's prediction. Using this framework across 11 attribution methods, six open-source LVLMs, and two output modes (direct answer and step-by-step reasoning), we find that existing attribution methods often fail to identify the evidence used by LVLMs. To address this failure, we propose MedFocus, a concept-based attribution method that localizes clinically meaningful anatomical regions via unbalanced optimal transport and measures their causal effect on model outputs through targeted interventions. MedFocus produces spatial, concept-level, and token-level attributions and substantially outperforms prior methods, taking a step toward more trustworthy attribution for medical LVLMs. Our data and code are available at https://github.com/gzxiong/medfocus/.

Vision-language alignment using chest X-rays and radiology reports has emerged as an advanced paradigm for zero-shot classification and grounding of chest X-ray findings. However, standard contrastive learning typically treats radiographs and reports from different patients simply as negative pairs. This assumption introduces noisy negatives, as different patients frequently exhibit similar findings. Such noisy negatives cause semantic ambiguity and degrade performance in zero-shot understanding tasks. To address this challenge, we propose CoNNS, a concept-guided noisy-negative suppression framework. To support the negative suppression mechanism, unlike previous methods that use raw reports or templatized texts, we construct a hierarchical concept ontology using large language models. The ontology structures 41 key clinical concepts by explicitly modeling presence, attributes (location and characteristics), and texts (evidential segment and presence statement). Leveraging this ontology, we implement a cross-patient pair relabeling strategy comprising three steps: (1) Fine-Grained Breakdown to categorize pairs based on finding presence; (2) Noisy Negative Filtering to resolve semantic conflicts by removing false negatives; and (3) Hard Negative Mining to identify subtle attribute discrepancies using a lightweight language model. Finally, we propose a Concept-Aware NCE loss to align visual features with text while suppressing the identified noisy negatives. Extensive experiments across multi-granularity zero-shot grounding tasks and five zero-shot classification datasets validate that CoNNS outperforms existing state-of-the-art models. The code is available at https://github.com/DopamineLcy/conns.

Deep learning methods have demonstrated promising results in predicting BI-RADS scores from mammography images. However, the interpretation of these images can vary, leading to discrepancies even among radiologists. Given the inherent complexity of mammograms, training classification models solely on image labels often yields limited performance. To address this challenge, we curated 2313 mammogram images and their corresponding captions from two mammography atlases. Our proposed approach employs a multi-modal model that uses a pretrained PubMedBERT as the language component. By training this model on image-text pairs with contrastive learning, we enable the vision encoder to absorb the rich information contained in the captions, thereby improving its understanding of mammography findings. We then fine-tune the vision encoder on two datasets for BI-RADS prediction, achieving superior performance compared with models trained without this pretraining, particularly when labeled samples are scarce. The improvement in the 3-class average F1 score ranges from +1% to +14%: a +1% increase with 40K training samples, and a +14% increase with 1K samples. Furthermore, our experiments reveal that 2K image-text pairs from mammography atlases can be more informative than 2K labeled samples for label prediction, with an average margin of +1.1% when more than 10K training samples are available. Overall, our work provides a vision-language model for mammography and highlights the value of textual information from mammography atlases. In addition, we publicly release preprocessed mammography images of the TEKNOFEST dataset. The training code, pre-trained model weights, data extraction scripts, and the released dataset are publicly available at: https://github.com/igulluk/MAM-CLIP

Medical vision-language models (VLMs) have rapidly advanced as general-purpose multimodal assistants, yet their deployment in 3D Computed Tomography (CT) analysis remains constrained by a persistent mismatch between optimization objectives and clinical rigor. Current Reinforcement Learning (RL) paradigms still rely on lexical proxy signals that induce \textit{Evaluation Hallucinations}'', where models optimize linguistic fluency rather than factual clinical correctness, leading to diagnostically critical errors. To bridge this gap, we introduce the \textbf{Clinical Abnormality Benchmarking Substrate (CABS)}, a structured system that decomposes radiology reports into verifiable clinical semantic units. Using CABS, we identify a \textit{Mechanistic Divergence}'' in standard RL, where surface-similarity rewards drive policy gradients to bypass medical facts. We therefore propose \textbf{Trajectory-Integral Feedback GRPO (TIF-GRPO)}, a novel framework integrating control-theoretic principles into policy optimization. By formulating clinical reasoning as a pseudo-temporal trajectory for anomaly discovery, TIF-GRPO regulates anatomy-aware rewards via an integral feedback loop that penalizes persistent omissions as cumulative state errors and suppresses hallucinations as excessive control effort. Experiments on 3D CT benchmarks demonstrate that our approach significantly enhances abnormality detection and clinical faithfulness, establishing a new paradigm for fine-grained regulation in medical VLMs. Our project is available at \href{https://github.com/ZJU4HealthCare/TIF-GRPO}{GitHub}.

Medical foundation models have achieved remarkable clinical performance, yet their robustness under real-world perturbations remains underexplored. We present a robustness benchmark comprising 40 perturbation types (12 base, 28 medical-specific) across eight imaging modalities, evaluating five VLMs (LLaVA-Med, MedGemma, MedGemma-1.5, Gemini-2.5-flash and GPT-4o-mini) on VQA, visual grounding, and captioning, alongside two segmentation models (MedSAM, SAM-Med2D) with five fine-tuning strategies. Our findings reveal: (1) Fine-tuning strategy dominates robustness, with LoRA exhibiting nearly double the degradation of full fine-tuning, while SAM-Med2D's Adapter offers favorable efficiency-robustness trade-off. (2) Medical-specific perturbations disproportionately damage segmentation, with 9 of 15 top corruptions being domain-specific. (3) LoRA-tuned visual grounding drops over 40 points, whereas zero-shot captioning remains stable (<7% drop). Zero-shot VQA shows model-dependent robustness--medical models drop under 20% while Gemini-2.5-flash drops 54%. General-purpose VLMs achieve higher VQA accuracy but fail on grounding; among medical VLMs, MedGemma demonstrates the best overall stability. These results provide deployment guidelines and underscore the necessity of domain-specific robustness evaluation for medical AI. Our code is available at: https://abnerai.github.io/MedFM-Robust.

Vision-language models (VLMs) are increasingly proposed for clinical decision support, yet their reliability in real-world scenarios that require integrating both visual and textual context from medical records remains poorly characterized. This paper identifies three failure modes: (1) modality over-reliance on text over images, (2) spurious reliance on irrelevant clinical history, and (3) prompt sensitivity across semantically equivalent inputs. We evaluate a diverse set of general-domain and medically-tuned open and closed VLMs on chest x-ray tasks using MIMIC-CXR. By systematically manipulating image-text alignment, clinical history, and prompt formulations, we found that VLM decisions are dominated by the text modality, even when visual evidence is available. Moreover, we observed that VLMs are heavily influenced by irrelevant reports, while minor prompt changes can reverse correct image-based predictions. Our findings underscore the need for explicit safeguards and stress-testing before considering the use of these models in clinical practice.

Recent progress in deep learning has significantly advanced CT image analysis, particularly for segmentation tasks. However, these advances are largely confined to image-level pattern recognition, with most methods lacking explicit anatomical or contextual reasoning. Large vision-language models introduce linguistic context into image analysis, yet most approaches typically focus on a single task, which is insufficient for clinical workflow analysis that requires multiple fine-grained types of analysis, such as anatomy detection and segmentation. In this paper, we propose a unified autoregressive framework that integrates language-guided visual reasoning into CT interpretation. Our method introduces task-routing tokens that trigger detection and segmentation heads conditioned on the hidden states of a large vision-language model, enabling coherent generation of visual outputs (e.g., masks and bounding boxes) and textual reasonings. To progressively enhance localisation accuracy and semantic clarity, we further design a "closer-look" mechanism that allows the model to perform progressive coarse-to-fine visits to regions of interest under refined fields of view. To support model training and evaluation, we curated a new multimodal CT dataset containing pixel-wise masks, bounding boxes, spatial prompts, and structured descriptions for visual objects constructed through an AI-assisted annotation process with human verification. Experiments on public benchmarks demonstrate consistent improvements over the SoTA, achieving up to 1.0% Dice on BTCV and 1.7% Dice on MosMed+, while additionally providing appearance reasoning outputs. The code and dataset will be available.

Vision-Language Models (VLMs) facilitate medical visual question answering (MedVQA) by jointly interpreting images and text. However, existing models typically depend on large architectures and closed-set answers, which limits their efficiency and potential clinical applicability. To overcome these shortcomings, we introduce RoiMAM, an efficient VLM. It integrates a training-free ROI Generation Module with Semantic Selective Suppression to focus on lesion-relevant regions, alongside a Text Prompt Enhancer module that provides modality-specific context without introducing training parameters. Compared to the widely used MedVInT-TD model, our design achieves efficient and accurate diagnosis at less than 20% of the model size, while improving accuracy by approximately 2% on SLAKE and 4.6% on PMC-VQA.

Concept Bottleneck Models (CBMs) offer interpretable alternatives to black-box predictors by introducing human-relatable concepts before the final output. However, existing CBMs struggle to verify whether predicted concepts correspond to the correct visual evidence, limiting their reliability. We propose a fine-grained CBM framework that grounds each concept in localized visual evidence, enabling direct inspection of where and how concepts are encoded. This design allows users to interpret predictions and verify that the model learns intended concepts rather than spurious correlations. Experiments on medical imaging benchmarks show that our learned concept space is information-complete and achieves predictive performance comparable to standard CBMs, while substantially improving transparency. Unlike post-hoc attribution methods, our framework validates both the presence and correctness of concept representations, bridging interpretability with verifiability. Our approach enhances the trustworthiness of CBMs and establishes a principled mechanism for human-model interaction at the concept level, paving the way toward more reliable and clinically actionable concept-based learning systems.

Fine-grained Vision-Language Pre-training (FVLP) demonstrates significant potential in 3D medical image understanding by aligning anatomy-level visual representations with corresponding textual descriptions. However, existing FVLP paradigms often suffer from severe representation collapse in the textual embedding space, where text embeddings of distinct anatomical structures become highly clustered and indistinguishable. This distributional degeneracy renders the model hypersensitive to prompt variations, hindering reliable clinical deployment. To address these challenges, we propose a novel Cross-Anatomy Global-Local Contrastive Learning framework (CA-GCL). CA-GCL introduces a global contrastive objective that enforces separation between anatomical categories in the latent space, effectively counteracting the aggregation tendency induced by local alignment. Furthermore, we incorporate a clinical-aware text augmentation strategy based on permutation invariance and partial completeness to enhance robustness against descriptive incompleteness. Extensive evaluations on the CT-RATE and Rad-ChestCT datasets show that CA-GCL achieves comparable zero-shot abnormality detection performance to existing VLP paradigms, while demonstrating substantially better robustness to prompt variations: on canonical templates it obtains higher mean AUC with lower variance, and on non-canonical templates it remains stable whereas baselines degrade markedly. These results validate CA-GCL as an effective framework for robust 3D medical image understanding.

Surgical scene understanding is a cornerstone of computer-assisted intervention. While recent advances, particularly in surgical image segmentation, have driven progress, real-world clinical applications require a more holistic understanding that jointly captures procedural context, semantic reasoning, and precise visual grounding. However, existing approaches typically address these components in isolation, leading to fragmented representations and limited semantic consistency. To address this limitation, we propose SurgMLLM, a unified surgical scene understanding framework that bridges high-level reasoning and low-level visual grounding within a single model. Given surgical videos, SurgMLLM fine-tunes a multimodal large language model (MLLM) to support structured interpretability reasoning, which is used to jointly model phases, instrument-verb-target (IVT) triplets, and triplet-entity segmentation tokens. These tokens are then temporally aggregated and serve as prompts for a segmentation network, enabling accurate pixel-wise grounding of triplet instruments and targets. The entire framework is trained end-to-end with a unified objective that couples language-based reasoning supervision with visual grounding losses, promoting coherent cross-task learning and clinically consistent scene representations. To facilitate unified evaluation, we introduce CholecT45-Scene, extending CholecT45 dataset with 64,299 frames of pixel-level mask annotations for instruments and targets, aligned with existing triplet labels. Extensive experiments show that SurgMLLM significantly advances surgical scene understanding, improving the primary triplet recognition metric AP_IVT from 40.7% to 46.0% and consistently outperforming prior methods in phase recognition and segmentation. These results highlight the effectiveness of unified reasoning-and-grounding for reliable, context-aware surgical assistance.

Chest radiograph interpretation requires temporal reasoning over prior and current studies, yet most vision-language models are trained on static image-report pairs and lack explicit supervision for modeling longitudinal change. We introduce CheXTemporal, a dataset for temporally grounded reasoning in chest radiography consisting of paired prior-current chest X-rays (CXR) with finding-level temporal and spatial annotations. The dataset includes a five-class progression taxonomy (new, worse, stable, improved, resolved), localized spatial supervision of pathology, explicit spatial-temporal alignment across paired studies, and multi-source coverage for cross-domain evaluation. We additionally construct a 280K-pair silver dataset with automatically derived temporal and anatomical supervision for large-scale evaluation under weaker supervision. Using these resources, we evaluate multiple state-of-the-art vision-language CXR models on grounding and progression-classification tasks in a zero-shot setting. Across both gold and silver evaluations, current models exhibit consistent limitations in spatial grounding, fine-grained temporal reasoning, and robustness under distribution shift. In particular, models perform substantially better on salient progression categories such as worse than on temporally subtle states such as stable and resolved, suggesting limited modeling of longitudinal disease evolution in chest radiography.

Self-verification, re-invoking the same vision language model (VLM) in a fresh context to check its own generated answer, is increasingly used as a default safety layer for medical visual question answering (VQA). We argue that this practice is fundamentally unreliable. We introduce [METHOD NAME], a diagnostic framework for mapping the reliability boundary of medical VLM self-verification by decomposing verifier behavior into discrimination capability and agreement bias. Because the verifier and answer generator are capacity-coupled, the verifier can overly agree with the generator, creating a verification mirage: a regime with both high verifier error and high agreement bias, driven by false acceptance of incorrect answers. Evaluating six open-weight VLMs across five medical VQA datasets and seven medical tasks, we find that this boundary is strongly task-conditioned. Knowledge-intensive clinical tasks fall deepest into the mirage, simpler tasks are more resistant, and perceptual tasks lie in between. Verification also fails to provide an independent safety signal: logistic mixed-effects analysis shows that verifier error and agreement bias become more likely when the generator is wrong, while saliency analyses show that verifiers under-attend to image evidence relative to generators, a phenomenon we call the lazy verifier. Cross-verification reduces but does not eliminate the mirage. Moreover, when verification is reused in multi-turn actor-verifier loops, most initially wrong answers become locked in by false verification. Since our experiments use clean benchmarks, the observed reliability boundary likely underestimates failures in real clinical deployment.

Large vision-language models (VLMs) demonstrate strong performance in medical image understanding, but frequently generate clinically plausible yet incorrect statements, raising significant safety concerns. Existing medical hallucination benchmarks primarily focus on 2D imaging with one-shot diagnostic questions, offering limited insight into whether predictions are grounded in correct localization and abnormality identification, allowing critical reasoning errors to remain hidden behind seemingly correct diagnoses. We introduce Med-StepBench, the first large-scale benchmark for step-wise hallucination detection in 3D oncological PET/CT, comprising over 12,000 images and more than 1,000,000 image-statement pairs across volumetric and multi-view 2D data, which decomposes clinical reasoning into four expert-designed diagnostic stages. Using clinician-verified annotations, we perform the first step-level evaluation of general-purpose and medical VLMs, revealing systematic failure modes obscured by aggregate accuracy metrics. Furthermore, we show that current VLMs are highly susceptible to adversarial yet clinically plausible intermediate explanations, which significantly amplify hallucinations despite contradictory visual evidence. Together, our findings highlight fundamental limitations in grounding multi-step clinical reasoning and establish Med-StepBench as a rigorous benchmark for developing safer and more reliable medical VLMs.

Plant disease diagnosis is critical for food security, yet training disease-recognition models that generalize across crops, pathogens, and field conditions remains challenging because labeled disease images are far less abundant and standardized than data for other biotic stresses such as insects or weeds. Frontier vision-language models offer new opportunities through improved visual reasoning, but they still struggle with fine-grained disease identification due to the lack of structured, crop-specific symptom knowledge. To address this gap, we curate the largest plant disease image--symptom dataset to date, covering 335 crops, 1{,}251 disease classes, and approximately 839K images, designed to support training-free, agentic disease prediction. A scalable automated pipeline generates source-grounded symptom descriptions in which each claim is linked to a verbatim web quote; domain experts validate sampled crops and reconcile disease-name variants across sources. As a baseline, we introduce an autonomous visual reasoning agent that identifies anatomical context, narrows candidate diseases using symptom knowledge, sequentially compares reference images, and produces a fully explainable reasoning trace. Incorporating symptom knowledge improves accuracy by 16.2 percentage points on average at the full reference budget, with consistent gains across all four evaluation crops. Because the framework only requires crop-specific reference images and symptom knowledge, it can be extended to new crops without retraining, while the agentic baseline can directly benefit from future improvements in foundation model capabilities. Dataset and code are available at:https://sage-dataset.github.io/.

Large-scale 3D vision-language models (VLMs) like LLaVA-3D offer strong spatial reasoning but are difficult to deploy due to high computational costs. We propose a knowledge distillation framework that transfers spatial reasoning from a 7B teacher to a 2.29B student model. Our approach achieves 8.7x lower inference latency and a 3x reduction in model size while retaining 54-72% of the teacher's performance. The framework utilizes VGGT as the vision encoder and a multi-task distillation pipeline with uncertainty-aware loss weighting. To improve reasoning without chain-of-thought (CoT) data, we introduce "Hidden CoT": learnable latent tokens that serve as an internal scratchpad before answer generation. This is the first use of latent scratchpad reasoning in distilled 3D VLMs. The student model jointly performs spatial description, depth estimation, and object detection. Experiments on ScanNet and 3D-FRONT show strong spatial understanding, reaching 68-72% accuracy on proximity and contact tasks. Our framework enables efficient 3D scene QA on resource-constrained platforms.

Medical vision-language models (VLMs) and AI agents have made significant progress in learning to analyze and reason about clinical images. However, existing medical visual question answering (VQA) benchmarks collapse model capabilities into a single accuracy score, obscuring where and why models fail. We propose DeepTumorVQA, a hierarchical benchmark that follows the multi-stage evidence chain in tumor diagnosis and decomposes 3D CT reasoning into four stages: recognition, measurement, visual reasoning, and medical reasoning. Higher-level questions remain independently scorable, while their ground-truth evidence chains are defined over lower-level primitives. The benchmark contains 476K questions across 42 clinical subtypes on 9,262 3D CT volumes. In addition to a direct reasoning mode for VLMs, DeepTumorVQA provides tool-interaction environments for agent evaluation, where a model can call external tools, including segmentation models, measurement programs, and medical knowledge modules, before answering the question. Evaluating over 30 model configurations, we find that reliable quantitative measurement is the primary bottleneck, making later-stage visual and medical reasoning harder for VLMs, while tool augmentation substantially mitigates this issue. When tools are available, leveraging medical knowledge and tools to reason about medical images becomes a new challenge. We further show that ground-truth step-by-step tool-use traces from DeepTumorVQA can supervise agents and reduce tool-use and reasoning failures. This stage-wise progression from recognition to measurement to visual and medical reasoning provides a concrete roadmap for future medical VLM and AI agent studies. All data and code are released at https://github.com/Schuture/DeepTumorVQA.

Vision--language models (VLMs) show promise for clinical decision support in radiology because they enable joint reasoning over radiological images and clinical text, thereby leveraging complementary clinical information. However, radiological findings are long-tailed in practice, leaving some conditions underrepresented and making zero-shot inference essential. Yet current CLIP-style medical VLMs are sensitive to prompt variations and often lack trustworthy external knowledge at inference time, which hinders reliable clinical deployment. We present \textit{KEPIL}, a prompt-robust framework that integrates curated medical knowledge to stabilize zero-shot generalization. KEPIL comprises: (i) \emph{dynamic prompt enrichment} using ontologies with LLM assistance, (ii) a \emph{semantic-aware contrastive loss} aligning embeddings of equivalent prompt variants via a dual-embedding objective, and (iii) \emph{entity-centric report standardization} to yield ontology-aligned representations. Across seven benchmarks, KEPIL achieves state-of-the-art zero-shot inference performance; under prompt-variation tests, it improves AUC by $6.37\%$ on \textit{CheXpert} and by $4.11\%$ on average. These results suggest that structured knowledge and robust prompt design are key to clinically reliable radiology-facing VLMs. Code will be released at https://github.com/Roypic/KEPIL.

Recent advances in 3D medical vision-language models have enabled joint reasoning over volumetric images and text, showing strong performance in medical visual question-answering (VQA) and report generation. Despite this progress, it remains unclear whether these models learn spatially grounded anatomy from 3D volumes or rely primarily on learned priors and language correlations. This uncertainty stems from the lack of systematic evaluation of semantic-spatial reasoning in volumetric medical VLMs for clinically reliable decision support. To address this gap, we introduce CT-SpatialVQA, a benchmark designed to evaluate semantic-spatial reasoning in 3D CT data. The benchmark comprises 9077 clinically grounded question-answer (QA) pairs derived directly from 1601 radiology reports and CT volumes, which are validated via a robust LLM-assisted pipeline with a 95% human consensus agreement rate. Our dataset requires explicit anatomical localization, laterality awareness, structural comparison, and 3D inter-structure relational reasoning. We also introduce a standardized evaluation protocol and benchmark eight 3D medical VLMs, finding severe degradation on semantic-spatial reasoning tasks, averaging 34% accuracy and often below random, highlighting the need for deeper integration of volumetric evidence for trustworthy clinical use.

Medical vision--language models (VLMs) are usually evaluated on intact image--question pairs, but trustworthy clinical use requires a stronger property: a model must recognise when the evidential basis for an answer has failed. We study this through silent failures under perturbed evidence, where a vision-required medical question is paired with a false premise, wording perturbation, knowledge-only rewrite, or ROI-corrupted image, yet the model returns a fluent non-refusal answer. We introduce medvigil, a 300-case evaluation suite drawn from four public medical VQA sources, supervised end to end by four board-certified radiologists: every gold answer, refusal option, candidate-answer set, paraphrase, false-premise trap, ROI box, and clinical risk tier is clinician-authored. Two attending radiologists annotate every case in parallel, a senior radiologist consolidates the released manifest, and a separate fourth radiologist independent of construction answers every probe to provide the human reference baseline. The release contains 2556 MCQ probes, 240 counterfactual triplets, physician-adjudicated risk-tier and answerability flags, ROI boxes, and a paired open-ended variant. We report seven correctness-conditioned audit metrics that summarise into the medvigil Composite Score (MCS), and audit 16 vision-capable models plus two text-only baselines. The independent radiologist scores MCS 83.3 at silent-failure rate 5.8%, leaving a 14.1-point composite headroom above the strongest audited model (Claude Opus 4.7 at 69.2). The benchmark and evaluation harness are publicly released.

When a chest X-ray shows consolidation but the question asks which finding is present, a medical vision-language model may answer "No consolidation." This is more than an incorrect choice: it is a polarity reversal that emits a clinical statement contradicting the image. We study this failure as negated-option attraction, where a model is drawn to a negated answer option even when it conflicts with both the visual evidence and the question. We introduce CXR-ContraBench (Chest X-Ray Contradiction Benchmark), a diagnostic benchmark spanning internal ReXVQA slices and external OpenI and CheXpert protocols. The benchmark centers on present-finding questions, where selecting "No X" despite visible X creates the main clinical risk, and uses absent-finding questions as secondary tests of whether models copy negated wording. Across CheXpert protocols, the failure is substantial and persistent. On a strict direct presence probe, MedGemma and Qwen2.5-VL reach only 31.49% and 30.21% accuracy, respectively; on a matched 135,754-record CheXpert training-split protocol, both models select negated options on over 62% of presence questions. Chain-of-thought prompting reduces some presence-side reversals but does not eliminate them and can amplify absence-side contradictions. Finally, QCCV-Neg (Question-Conditioned Consistency Verifier for Negation) deterministically repairs the measured polarity-confused subset without retraining, raising MedGemma and Qwen2.5-VL to 96.60% and 95.32% accuracy on the direct presence probe. These results show that standard accuracy can hide a clinically meaningful inference-time polarity failure. Source code and benchmark construction scripts are available at https://github.com/fangzr/cxr-contrabench-code.

Zero-shot anomaly localisation via vision-language models (VLMs) offers a compelling approach for rare pathology detection, yet its performance is fundamentally limited by the absence of healthy anatomical context. We reformulate zero-shot localisation as a comparative inference problem in which anomalies are identified through structured comparison against reference distributions of normal anatomy. We introduce WALDO, a training-free framework grounded in optimal transport theory that enables comparative reasoning through: (i) entropy-weighted Sliced Wasserstein distances for anatomically-aware reference selection from DINOv2 patch distributions, (ii) Goldilocks zone sampling exploiting the non-monotonic relationship between reference similarity and localisation accuracy, and (iii) self-consistency aggregation via weighted non-maximum suppression. We theoretically analyse the Goldilocks effect through distributional divergence, and show that references with moderate similarity minimize a bias-variance trade-off in comparative visual reasoning. On the NOVA brain MRI benchmark, WALDO with Qwen2.5-VL-72B achieves $43.5_{\pm1.6}\%$ mAP@30 (95% CI: [40.4, 46.7]), representing a 19% relative improvement over zero-shot baselines. Cross-model evaluation shows consistent gains: GPT-4o achieves $32.0_{\pm6.5}\%$ and Qwen3-VL-32B achieves $32.0_{\pm6.6}\%$ mAP@30. Paired McNemar tests confirm statistical significance ($p<0.01$). Source code is available at https://github.com/bkainz/WALDO_MICCAI26_demo .

Vision-language models hold considerable promise for ophthalmology, but it remains unclear which training data source best conveys expert domain knowledge. Existing ophthalmic models are trained on fixed text templates, medical reports, or general biomedical literature, sources that have never been compared under matched conditions. To include domain-specific literature in this comparison, we present PubMed-Ophtha, a hierarchical dataset with high domain density of 102,023 panels with their subcaptions from 15,842 open-access articles in PubMed Central. We then finetuned identical CLIP models on each source, using a general biomedical literature model as baseline, and found that domain-specific literature achieved the best average performance across 110 clinical tasks, reaching a mean linear probing AUROC of 88.63% ahead of medical reports (85.68%). Restricting the dataset to fundus images, to the image count of the medical report dataset, or to articles unrelated to the evaluation datasets did not reduce performance, indicating that the gains likely stem from domain density. We release the dataset, the finetuned models, and the full generation pipeline.

Vision-language models (VLMs) have shown potential for automated radiology report generation, yet existing approaches rely on global embedding compression of volumetric data, often leading to hallucinated findings and limited anatomical grounding in 3D CT imaging. We introduce MedScribe, a hypothesis-driven framework that reformulates report generation as an iterative evidence acquisition process rather than a single-pass encoding task. MedScribe models reporting as a sequential decision process in which a large language model dynamically invokes pathology-specific diagnostic tools to extract localized volumetric features. These structured features are used to query a multidimensional retrieval space aligned with pathology-specific textual evidence. By explicitly accumulating quantitative evidence prior to synthesis, the framework enforces fine-grained grounding and reduces unsupported claims. Without task-specific fine-tuning, MedScribe improves clinical accuracy, factual consistency, and interpretability on CT-RATE and RadChestCT compared to state-of-the-art 2D and 3D VLMs, demonstrating the value of hypothesis-driven reasoning for reliable medical image reporting.

Zero-shot vision-language models (VLMs) offer a promptable alternative to task-specific training for gross tumor volume (GTV) delineation in non-small-cell lung cancer (NSCLC), but the prompt dimensions that govern their spatial behavior remain poorly understood. We study this question by probing alignment directions in VoxTell on a held-out internal NSCLC tumor dataset through sub-prompt decomposition into diagnosis, demographic, staging, anatomical, generic, and irrelevant controls; attribute-wise perturbation robustness; specificity ladders; and cross-case prompt swaps, while benchmarking against fine-tuned and zero-shot baselines using the Dice Similarity Coefficient (DSC) with Wilcoxon signed-rank tests and Benjamini-Hochberg correction. Alignment analyses revealed that anatomical location is the dominant driver of VoxTell's spatial attention: 63.4 percent of location perturbations caused catastrophic drops, prompt specificity improved from generic to full descriptions except for diagnosis-only prompts, irrelevant prompts correctly yielded zero segmentation, and cross-case prompt swaps confirmed patient-specific conditioning (matched DSC 0.906 vs. mismatched 0.406). Histology and stage substitutions had minimal effect, indicating that the model prioritizes "where to look" over "what to look for." In this context, VoxTell, operating fully zero-shot, achieved a mean DSC of 0.613, statistically indistinguishable from nnUNet (0.690, adjusted p = 0.156) and Ahmed et al. (0.675, adjusted p = 0.679), while significantly outperforming all other zero-shot models. Together, these findings argue that segmentation VLMs should be evaluated not only by Dice, but also by the prompt dimensions to which they align.

Large language models (LLMs) show promise in radiology but their deployment is limited by computational requirements that preclude use in resource-constrained clinical environments. We investigate whether small language models (SLMs) of 3-4 billion parameters can achieve strong multi-task radiology performance through LoRA fine-tuning, enabling deployment on consumer-grade CPUs. We train Qwen2.5-3B-Instruct and Qwen3-4B on 162K samples spanning 9 radiology tasks - RADS classification across 10 systems, impression generation, temporal comparison, radiology NLI, NER, abnormality detection, N/M staging, and radiology Q&A - compiled from 12 public datasets. Both models are evaluated on up to 500 held-out test samples per task with standardized metrics. Our key findings are: (1) LoRA fine-tuning dramatically improves performance over zero-shot baselines (RADS accuracy +53%, NLI +60%, N-staging +89%); (2) the two models exhibit complementary strengths - Qwen2.5 excels at structured generation tasks while Qwen3 dominates extractive tasks; (3) a task-outed oracle ensemble combining both models achieves the best performance across all tasks; (4) few-shot prompting with fine-tuned models hurts performance, demonstrating that LoRA adaptation is more effective than in-context learning for specialized domains; and (5) models can be quantized to GGUF format (~1.8-2.4GB) for CPU deployment at 4-8 tokens/second on consumer hardware. Our work demonstrates that small, efficiently fine-tuned models - which we collectively call RadLite - can serve as practical multi-task radiology AI assistants deployable entirely on consumer hardware without GPU requirements. Code and models are available at https://github.com/RadioX-Labs/RadLite

Chest X-ray interpretation is one of the most frequently performed diagnostic tasks in medicine and a primary target for AI development, yet current vision-language models are primarily trained on datasets of paired images and reports, not the cognitive processes and visual attention that underlie clinical reasoning. Here, we present CheXthought, a global, multimodal resource containing 103,592 chain-of-thought reasoning traces and 6,609,082 synchronized visual attention annotations across 50,312 multi-read chest X-rays from 501 radiologists in 71 countries. Our analysis reveals clinical reasoning patterns in how experts deploy distinct visual search strategies, integrate clinical context, and communicate uncertainty. We demonstrate the clinical utility of CheXthought across four dimensions. First, CheXthought reasoning significantly outperforms state-of-the-art vision-language model chain-of-thought in factual accuracy and spatial grounding. Second, visual attention data used as an inference-time hint recovers missed findings and significantly reduces hallucinations. Third, vision-language models trained on CheXthought data achieve significantly stronger pathology classification, visual faithfulness, temporal reasoning and uncertainty communication. Fourth, leveraging CheXthought's multi-reader annotations, we predict both human-human and human-AI disagreement directly from an image, enabling transparent communication of case difficulty, uncertainty and model reliability. These findings establish CheXthought as a resource for advancing multimodal clinical reasoning and the development of more transparent, interpretable vision-language models.

Generative vision-language models can produce fluent medical image captions but remain prone to hallucination, over-specific diagnostic claims, and factual inconsistency-serious issues in pathology. We investigate retrieval-guided generation (RGG) as a safer alternative, where captions are formed by summarizing expert text from visually similar cases rather than generated de novo. On the ARCH histopathology dataset, RGG improves semantic alignment with ground truth, achieving cosine similarity of $\approx$0.60 versus $\approx$0.47 from MedGemma, with non-overlapping confidence intervals indicating a robust gain. A pathologist-led qualitative review shows better preservation of morphology-relevant terminology and fewer unsupported diagnoses, while revealing failure modes such as concept mixing and inherited over-specific labeling. Overall, retrieval-guided captioning offers a more transparent and reliable approach with clearer opportunities for auditing than fully generative methods.