The paper introduces SPAR‑Bench, a set of eight probes designed to test whether medical vision models can reason about anatomy in abdominal CT scans. Experiments across five architectures and three foundation models—both frozen and fine‑tuned—show that while models can recall canonical organ locations, they fail to perform relational reasoning or spatial comparisons within a patient, even under zero‑shot transfer. The study also demonstrates that pooled probing underestimates a model’s relational capabilities and that open‑weight multimodal large language models perform poorly on these tasks.
arXiv:2607. 20993v1 Announce Type: cross Abstract: Large vision-language models are becoming increasingly dominant in 3D medical image interpretation, but we rarely know which internal units encode clinical findings or where that information lives in the representation.
arXiv:2608.21140v1 Announce Type: cross Abstract: Reliable spatial understanding is an important prerequisite for future medical vision-language systems that aim to support radiological report genera...
arXiv:2607. 25164v1 Announce Type: cross Abstract: A CT examination captures multiple organs, but many biomedical questions concern abnormalities, prognosis, or longitudinal change in a specific organ.
DALE-CT introduces depth‑aware 2D slice encoders that learn an anatomical world model of chest CT scans without 3D or positional supervision. By sampling self‑supervised views across a physical $z$‑axis slab, the encoder captures how anatomy changes between neighboring slices, enabling it to recover slice ordering and distinguish slices by anatomy alone. The model, trained on a large 287k‑scan corpus, achieves state‑of‑the‑art performance on CT‑RATE and is released with full code and evaluation tools.
arXiv:2610.06896v1 Announce Type: new Abstract: Frontier multimodal large language models (MLLMs) are increasingly positioned as general purpose visual reasoners as part of the quest for artificial g...
arXiv:2608.00231v2 Announce Type: replace Abstract: Volumetric CT vision-language pretraining learns 3D representations from scan-report pairs, but global and anatomy-aware objectives supervise only...
arXiv:2608. 05960v1 Announce Type: cross Abstract: Routine CT interpretation is inherently comprehensive, capturing incidental findings across the entire scan volume.
The paper introduces MedREAL, a unified framework that aligns linguistic reasoning with spatial grounding for medical visual question answering and segmentation. MedREAL employs Seg Anchored Reasoning Pooling (SARP) to extract semantic evidence from segmentation tokens and a Reasoning-to-Visual (R2V) fusion mechanism to integrate these features into a segmentation pipeline. Using the newly created MedRAVS-13K dataset, MedREAL achieves superior performance, reporting 68.49% gIoU and 70.47% cIoU, and generates evidence masks that consistently match textual diagnoses.
arXiv:2509.22404v2 Announce Type: replace Abstract: Anatomical understanding, which is the ability to identify, localize, or segment anatomical structures, is critical in medical image analysis; howe...
arXiv:2607. 22771v1 Announce Type: cross Abstract: Picking the frozen image encoder for a 3D~CT vision--language model (VLM), together with the token-compression scheme on top of it, is a search over many candidates.
NV-Reason-CT is a generative vision‑language model designed for chest and abdominal CT analysis that preserves native 3D visual encoding and incorporates radiologist‑guided reasoning. The system couples a 3D vision transformer with a language model, feeding all visual tokens and their 3D coordinates directly into language decoding to maintain volumetric spatial information. Trained on a curated corpus of about 550,000 multimodal instruction examples, the model supports abnormality classification, report generation, and interactive reasoning, achieving strong performance on CT benchmarks and reducing expert interpretation time by 50%.
arXiv:2603. 18846v3 Announce Type: replace-cross Abstract: Foundation models are used to extract transferable representations from large amounts of unlabeled data, typically via self-supervised learning (SSL).