Computer vision–enhanced signal-to-background analysis of indocyanine green near-infrared fluorescence to identify extrahepatic biliary anatomy during laparoscopic cholecystectomy

By
Philip J. Seger
Shinil Shah
Akhil Gorugantu
Nan Xi
Brendan T. Fox
Rishikesh Bhyri
Kaity Tung
Rita Glazer
Brian Quaranto
Junsong Yuan
Erik Wilson
Gene Yang
Steven D. Schwaitzberg
Peter C. W. Kim
October 8, 2026

Surgical Endoscopy

Overview

The article context describes the rationale for using computer vision (CV) to improve interpretation of indocyanine green near-infrared fluorescence (ICG-NIRF) during laparoscopic cholecystectomy (LC). ICG-NIRF can improve extrahepatic bile duct visualization, but hepatic background fluorescence, heterogeneous protocols, and subjective interpretation remain limitations; the supplied material does not provide the study’s methods or results.

Background

LC is performed frequently, and bile duct injury (BDI), although uncommon, is associated with substantial morbidity and long-term quality-of-life effects. The critical view of safety (CVS) is the preferred approach for ductal identification, but inflammation, anatomic variation, and adipose tissue can make it difficult to obtain. Intraoperative cholangiography can reduce BDI but adds operative time, workflow disruption, and radiation exposure; ICG-NIRF offers radiation-free, real-time biliary visualization. Its utility can be constrained by background fluorescence and operator-dependent interpretation, motivating investigation of CV-based image analysis.

Data Highlights

Measure or context	Reported information
Annual LC procedures in the United States	As many as 1 million
Reported BDI incidence during LC	0.3–0.8%
Reported increase in morbidity at 1 year after BDI	Threefold
Reported annual U.S. cost associated with BDI	Approximately $1 billion
ICG-NIRF biliary visualization in systematic reviews cited in the context	Rates above 90%
Computer-vision study results	Not provided in the supplied article context

Key Findings

The supplied article context identifies CVS as the preferred approach for accurate ductal identification and specifies its three components: clearing the hepatocystic triangle, separating the lower gallbladder from the liver to expose the cystic plate, and confirming only two structures enter the gallbladder.
ICG-NIRF is described as a radiation-free method for real-time biliary mapping, with prior studies reporting improved extrahepatic bile duct visualization compared with white-light imaging.
Systematic reviews cited in the context report biliary visualization rates above 90%; the context states that ICG-NIRF may improve, but does not eliminate, BDI rates.
First-pass hepatic uptake of ICG can produce background fluorescence that limits recognition of biliary structures within the operative field.
The context reports that ICG dose and administration timing have been investigated, but approaches are heterogeneous and standardized protocols and recommendations are lacking.
CV segmentation models can extract pixel-wise features from surgical images; the supplied material presents this as a potential means of more objective scene interpretation but provides no results from the titled study.

Clinical Implications

The supplied context presents ICG-NIRF as an adjunct to anatomic identification, while retaining CVS as the preferred ductal-identification approach. It describes CV analysis as an investigational means of addressing image-interpretation limitations; no performance results or clinical recommendations from the titled study are included in the provided material.

Conclusion

ICG-NIRF can enhance biliary visualization during LC, but background fluorescence and variable interpretation remain limitations. The supplied material motivates CV-enhanced analysis but does not report whether it improves operative or patient outcomes.