Metraipolis › archive › dc_2fbdd9332da4

Do Medical Vision Models Reason About Anatomy? Probing the Spatial Inductive Biases of Learned Visual Representations — arXiv:2608.28092

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.28092
License (as archived)
CC0-1.0
Archived
2026-10-17T10:48:05.856Z
SHA-256 of the archived text
22f1c84fa328e150b12af4515dd8dd9814c270906c9d29cf34ee089bab336d05
Stable URL of this page
https://metraipolis.com/documento/dc_2fbdd9332da4

How to cite

Do Medical Vision Models Reason About Anatomy? Probing the Spatial Inductive Biases of Learned Visual Representations — arXiv:2608.28092 — arXiv (en), https://arxiv.org/abs/2608.28092 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_2fbdd9332da4 (Metraipolis archive dc_2fbdd9332da4, sha256 22f1c84fa328e150b12af4515dd8dd9814c270906c9d29cf34ee089bab336d05)
BibTeX
@misc{dc2fbdd9332da4,
  title = {Do Medical Vision Models Reason About Anatomy? Probing the Spatial Inductive Biases of Learned Visual Representations — arXiv:2608.28092},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.28092},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_2fbdd9332da4; sha256 22f1c84fa328e150b12af4515dd8dd9814c270906c9d29cf34ee089bab336d05},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.28092. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Interpreting a CT scan means comparing structures on either side, judging how far apart organs sit, and knowing where each one belongs. Medical vision encoders are evaluated on diagnostic accuracy, or through assembled multimodal systems where a failure is hard to attribute, so it remains unclear whether their representations support any of this. We construct SPAR-Bench, eight probes over multi-organ abdominal CT that separate coordinate localization, relational reasoning, and spatial queries, and apply them to five architectural configurations and three medical foundation models, frozen and finetuned. Probes that ask for a comparison within the slice stay at chance, and neither pretraining scale, finetuning, nor architecture closes the gap. Probes that appear solved in domain fall to chance under zero-shot transfer, indicating that their accuracy reflects recall of canonical anatomy rather than computation over the image. Reading the same frozen features with a pooled head rather than the full set of tokens moves relational recovery from 0.7% to 67.8%, so pooled probing understates what a representation holds. Questions the encoders answer well are answered at chance by four open-weight MLLMs. Our results suggest these encoders carry a map of where organs usually lie, and little of the machinery for comparing structures within a particular patient. Code and data will be available at this https URL .

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-31 in cs.AI).
Categorie arXiv: Image and Video Processing (eess.IV) ; Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2608.28092 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-17T10:33:42.495Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-31?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).