Metraipolis › archive › dc_963f444233dc

ModaLens: Measuring Image Sensitivity in Report-Conditioned Medical VLMs — arXiv:2609.15635

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.15635
License (as archived)
CC0-1.0
Archived
2026-10-15T22:53:56.574Z
SHA-256 of the archived text
90e1676f6acbdd8327e0b1083088b8cf484b93d1a6e4016f76ac02aa13ebeb0a
Stable URL of this page
https://metraipolis.com/documento/dc_963f444233dc

How to cite

ModaLens: Measuring Image Sensitivity in Report-Conditioned Medical VLMs — arXiv:2609.15635 — arXiv (en), https://arxiv.org/abs/2609.15635 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_963f444233dc (Metraipolis archive dc_963f444233dc, sha256 90e1676f6acbdd8327e0b1083088b8cf484b93d1a6e4016f76ac02aa13ebeb0a)
BibTeX
@misc{dc963f444233dc,
  title = {ModaLens: Measuring Image Sensitivity in Report-Conditioned Medical VLMs — arXiv:2609.15635},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.15635},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_963f444233dc; sha256 90e1676f6acbdd8327e0b1083088b8cf484b93d1a6e4016f76ac02aa13ebeb0a},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.15635. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

A radiology report can already answer a clinical question, so it is hard to tell whether a vision-language model also uses the image. ModaLens, a paired image-swap audit, measures how report availability changes image sensitivity: MedGemma-27B on 3,199 paired MIMIC-CXR cases from 293 patients, all 14 questions per case (13 finding-specific and one composite), each image replaced by one from another study, usually of the same patient, with question and report fixed. Under an explicit answer instruction, the model's generated answer changes on 4.26 percent of trials with the report and 20.94 percent without it, a paired increase of 16.7 points (patient-clustered 95 percent CI 15.6 to 17.7), so report availability reduces image-swap sensitivity under this protocol; the original prompt with a lowercase first-token readout gives 4.70 percent against 17.07 percent, and substitutions also move continuous answer scores where the binary prediction does not change. The labels are derived from reports, which limits conclusions about visual correctness; the direction replicates in two further model lineages. Code, the exact prompts and a run record for every number are at this https URL .

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-15 in cs.AI).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.15635 — 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 2026-10-15T22:29:03.899Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-15?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).