Metraipolis › archive › dc_2f52243ba60b

NV-Reason-CT: 3D Visual Language Model for CT Analysis — arXiv:2609.27511

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.27511
License (as archived)
CC0-1.0
Archived
2026-10-08T09:12:42.382Z
SHA-256 of the archived text
d1019fd6d13ea4ca84aa1f3bc01bbbcdcbad7098a62e48db59810918327fbd93
Stable URL of this page
https://metraipolis.com/documento/dc_2f52243ba60b

How to cite

NV-Reason-CT: 3D Visual Language Model for CT Analysis — arXiv:2609.27511 — arXiv (en), https://arxiv.org/abs/2609.27511 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_2f52243ba60b (Metraipolis archive dc_2f52243ba60b, sha256 d1019fd6d13ea4ca84aa1f3bc01bbbcdcbad7098a62e48db59810918327fbd93)
BibTeX
@misc{dc2f52243ba60b,
  title = {NV-Reason-CT: 3D Visual Language Model for CT Analysis — arXiv:2609.27511},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.27511},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_2f52243ba60b; sha256 d1019fd6d13ea4ca84aa1f3bc01bbbcdcbad7098a62e48db59810918327fbd93},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.27511. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

We present NV-Reason-CT, a generative vision--language model for chest and abdominal CT combining native 3D visual encoding with radiologist-guided reasoning. The model couples a native 3D vision transformer with a language model, passing all visual tokens and their explicit 3D coordinates into language decoding without further spatial token merging. This retains volumetric spatial information within the vision encoder and through the language model's positional encoding during joint processing with text. We train on a curated corpus of approximately 550,000 multimodal instruction examples from 70,111 unique CT image inputs, combining standardized reports, abnormality-focused and anatomy-specific questions, multi-turn interactions, and radiologist-authored reasoning from recorded and transcribed expert CT interpretations. Expert annotations provide direct supervision and guide additional report-grounded synthetic reasoning. End-to-end supervised fine-tuning (SFT) is followed by Group Relative Policy Optimization (GRPO), with verifiable rewards over chest and abdominal abnormality sets. The model supports abnormality classification, report generation, and interactive reasoning with reviewable observations, differential diagnoses, and uncertainty. Evaluation spans public CT benchmarks and a held-out NIH cohort. On CT-RATE, NV-Reason-CT achieves a macro-F1 of 0.614 and macro-AUROC of 0.871 without a task-specific classification head; generated reports achieve a report-derived macro-F1 of 0.592. In a preliminary study with expert radiologists, AI-assisted review received favorable confidence ratings and was associated with a 50% reduction in average reported interpretation and reporting time. We release the model and training code to support reproducible research on explainable AI for volumetric medical imaging.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-24 in cs.AI).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.27511 — 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 2026-10-08T08:13:26.685Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-24?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).