Skip to content
DraconiticResearch
Search Sources API
197,688 studies31 sourcesUpdated 20 min ago
Peer-reviewedCC BY-NC-NDnpj Digital Medicine

Quantifying the plausibility gap in generative AI for surgical video generation with expert assessment

  • Zhen Chen1,2
  • Qing Xu3
  • Jinlin Wu4
  • Biao Yang5
  • Yuhao Zhai6
  • Geng Guo5
  • Jing Zhang6
  • Yinlu Ding6
  • Nassir Navab7
  • Jiebo Luo8
8 affiliations
  1. 1Department of Data Science and Artificial Intelligence, The Hong Kong Polytechnic University, Hong Kong, China
  2. 2Research Institute for Smart Ageing, The Hong Kong Polytechnic University, Hong Kong, China
  3. 3Department of Computer Science, University of Nottingham, Nottingham, UK
  4. 4Institute of Automation, Chinese Academy of Sciences, Beijing, China
  5. 5Department of Neurosurgery, The First Hospital, Shanxi Medical University, Taiyuan, China
  6. 6Department of Gastrointestinal Surgery, The Second Qilu Hospital, Shandong University, Jinan, China
  7. 7The Chair for Computer Aided Medical Procedures and Augmented Reality (CAMP), Technical University of Munich, Munich, Germany
  8. 8Department of Computer Science, University of Rochester, Rochester, NY, USA
Open at sourcePDFDOICite

Abstract

Generative artificial intelligence (AI) shows promise for physical simulation, yet its clinical application remains critically unexplored. We present SurgVeo, a pioneering expert-guided benchmark paired with the Surgical Plausibility Pyramid framework. Evaluating the Veo-3 and Wan2.2 models reveals a consistent plausibility gap: despite high visual fidelity, the majority of errors are surgical logic failures rather than visual quality deficits. By exposing this reliance on visual synthesis, this study provides a roadmap for developing surgical world models.

Related

  1. Challenges in Quantifying Upper Extremity Dysmetria in the Scale for the Assessment and Rating of Ataxia (SARA)
    D. Apaydin, J. D. Schmahmann
    medRxivPreprintAll rights reservedFull textNeurologymedRxiv
  2. Generative AI for Synthetic Medical Image Generation in Oncology: Addressing Data Scarcity in AI-Driven Cancer Diagnosis
    Ohmini Krishnamurthy Rajendran
    International Journal of Drug Delivery TechnologyPeer-reviewedLicence not statedFull text20 citedEpidemiology & public healthOpenAlex
  3. Radiographic union rate and predictors of nonunion after primary surgical implant generation network intramedullary nailing for open tibial shaft fractures at a major trauma center in Ethiopia
    Solomon Almaw Enquahone, Ashenafi Negash Tekle, Samuel Workineh Assefa +4
    OTA international : the open access journal of orthopaedic traumaPeer-reviewedCC BY-NC-NDFull textEpidemiology & public healthSurgery & anaesthesiaEurope PMC
  4. Quantifying the diagnostic time gap in precision oncology: A retrospective analysis of potential time savings from same-day cartridge-based molecular testing versus next-generation sequencing in cytology specimens
    Kevin Tran, Eveline Chen, W. Fraser Symmans +1
    PLOS OnePeer-reviewedCC BY 4.0Full textOncologyPLOS
  5. Case report: surgical total cardiac denervation for refractory vasospastic angina: first report with pre- and post-operative autonomic assessment
    European heart journal. Case reportsPeer-reviewedCC BYFull textCardiovascular medicineEurope PMC

Cite

BibTeXRIS
BibTeX
@article{chen2026quantifying,
  title = {{Quantifying the plausibility gap in generative AI for surgical video generation with expert assessment}},
  author = {Zhen Chen and Qing Xu and Jinlin Wu and Biao Yang and Yuhao Zhai and Geng Guo and Jing Zhang and Yinlu Ding and Nassir Navab and Jiebo Luo},
  journal = {npj Digital Medicine},
  year = {2026},
  month = sep,
  doi = {10.1038/s41746-026-03276-z},
  url = {https://doi.org/10.1038/s41746-026-03276-z},
  note = {CC BY-NC-ND},
}
RIS
TY  - JOUR
TI  - Quantifying the plausibility gap in generative AI for surgical video generation with expert assessment
AU  - Zhen Chen
AU  - Qing Xu
AU  - Jinlin Wu
AU  - Biao Yang
AU  - Yuhao Zhai
AU  - Geng Guo
AU  - Jing Zhang
AU  - Yinlu Ding
AU  - Nassir Navab
AU  - Jiebo Luo
PY  - 2026
DA  - 2026/09/26
JO  - npj Digital Medicine
DO  - 10.1038/s41746-026-03276-z
UR  - https://doi.org/10.1038/s41746-026-03276-z
AB  - Generative artificial intelligence (AI) shows promise for physical simulation, yet its clinical application remains critically unexplored. We present SurgVeo, a pioneering expert-guided benchmark paired with the Surgical Plausibility Pyramid framework. Evaluating the Veo-3 and Wan2.2 models reveals a consistent plausibility gap: despite high visual fidelity, the majority of errors are surgical logic failures rather than visual quality deficits. By exposing this reliance on visual synthesis, this study provides a roadmap for developing surgical world models.
SN  - 2398-6352
ER  -