Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Video

Updated: May 12, 2026

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
03:14

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

Sum-of-Checks: structured reasoning for surgical safety with large vision-language models.

Weiqiu You1, Cassandra Goldberg2, Amin Madani3,4

  • 1Department of Computer and Information Science, University of Pennsylvania, Philadelphia, PA, USA. weiqiuy@seas.upenn.edu.

International Journal of Computer Assisted Radiology and Surgery
|May 11, 2026
PubMed
Summary

Related Concept Videos

Deductive Reasoning01:16

Deductive Reasoning

Deductive reasoning, or deduction, is the type of logic used in hypothesis-based science. In deductive reasoning, the pattern of thinking moves in the opposite direction as compared to inductive reasoning, which means that it uses a general principle or law to predict specific results. From those general principles, a scientist can deduce and predict the specific results that would be valid as long as the general principles are valid.
For example, a researcher can deduce specific predictions...

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

The Role of Particulate Matter in Surgical Site Infection Pathogenesis: A Scoping Review.

The Journal of hospital infection·2026
Same author

Building Capacity for Surgical Artificial Intelligence Research.

Journal of the American College of Surgeons·2026
Same author

Regulating digital surgery to balance safety and innovation: a SAGES white paper.

Surgical endoscopy·2026
Same author

What are the ethical, legal and social considerations for recording, sharing and utilizing surgical videos? A thematic analysis and framework to guide stakeholders.

Surgical endoscopy·2026
Same author

Risk and liability in the deployment of AI systems for surgery: a SAGES white paper.

Surgical endoscopy·2026
Same author

Preventing Cat-Astrophe: Toxoplasma Seroprevalence and Prophylaxis Adherence After Allogeneic Stem Cell Transplant.

Transplant infectious disease : an official journal of the Transplantation Society·2026

A new Sum-of-Checks framework enhances the accuracy and transparency of AI assessment for the Critical View of Safety (CVS) in laparoscopic surgery. This method improves reliability for preventing bile duct injuries during cholecystectomy.

Area of Science:

  • Surgical Safety and Artificial Intelligence
  • Medical Imaging and Computer Vision

Background:

  • Bile duct injury (BDI) is a severe complication of laparoscopic cholecystectomy, necessitating accurate intraoperative safety assessments.
  • Large vision-language models (LVLMs) show promise for surgical analysis but lack auditability and reliability in safety-critical applications.

Purpose of the Study:

  • To develop and evaluate a novel framework, Sum-of-Checks, for improving the accuracy and auditability of LVLM-based assessment of the Critical View of Safety (CVS).
  • To enhance the reliability of AI systems in preventing bile duct injuries during laparoscopic cholecystectomy.

Main Methods:

  • The Sum-of-Checks framework decomposes CVS criteria into expert-defined reasoning checks, with LVLMs providing binary judgments and justifications for each check.
  • Evaluated on the Endoscapes2023 benchmark using three leading LVLMs, comparing Sum-of-Checks against direct prompting and chain-of-thought methods.
Keywords:
Critical View of SafetyLaparoscopic cholecystectomyLarge vision-language modelsSurgical reasoning

Related Experiment Videos

Last Updated: May 12, 2026

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
03:14

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

Main Results:

  • Sum-of-Checks demonstrated a 12-14% relative improvement in average frame-level mean average precision across all tested LVLMs and CVS criteria.
  • LVLMs exhibited reliability on observational checks but significant variability on decision-critical anatomical evidence assessments.

Conclusions:

  • Structuring surgical reasoning into expert-aligned verification checks enhances both accuracy and transparency in LVLM-based CVS assessment.
  • Separating evidence elicitation from decision-making is crucial for developing reliable and auditable surgical AI systems.