Analysis and mitigation of equipment-induced shortcuts in AI models for laparoscopic cholecystectomy
- 1Department of Computer Science, University of Toronto, Toronto, Ontario, Canada
- 2Vector Institute for Artificial Intelligence, Toronto, Ontario, Canada
- 3University Health Network, Toronto, Ontario, Canada
Abstract
Machine learning models have seen a lot of success in medical image segmentation domain. However, one of the challenges that they face are confounders or shortcuts: spurious correlations or biases in the training data that affect the resulting models. One example of such confounders for surgical machine learning is the setup of surgical equipment, including tools and lighting. Using the task of identification of safe and dangerous zones of dissection in laparoscopic cholecystectomy images and videos as a use-case, we inspect two equipment-induced biases: the location of surgical tools in the field of view and the direction of lighting, both of which are tightly related to the region of interest. We propose methods for evaluating the severity of these biases based on measurements of model consistency under real or simulated exposure to these shortcuts, and augmentation-based methods for mitigating them. We show that our tool bias mitigations based on pasting tool images in random locations improve the models’ prediction consistency under tool movements by 9 percentage points in the most inconsistent cases, and by 4 percentage points on average. Our lighting bias mitigations based on simulated lighting adjustments help reduce fraction of pixels originally predicted as belonging to the dangerous zone that may flip to safe under light changes from 5% to 1.5%, without compromising segmentation quality. Author summary: Artificial intelligence is increasingly used to support surgeons, for example by highlighting which tissue is safe to cut and which is dangerous during gallbladder removal (laparoscopic cholecystectomy), one of the most common abdominal operations. These systems can, however, learn the wrong lesson. We found that AI models for this task come to rely on where the surgeon places the instruments and where the camera light is pointed, rather than on the anatomy itself. Experienced surgeons tend to position their tools and lighting in predictable places, so the model takes a shortcut. The danger is that a less experienced surgeon, who handles equipment differently, could receive misleading guidance. This flaw is hard to detect because the pattern runs through all of the data, so ordinary testing does not reveal it. We developed a way to measure how strongly a model leans on these shortcuts, and a simple training method that breaks them by inserting instruments and extra lighting at random positions. Our approach makes predictions substantially more consistent without sacrificing accuracy, a step toward surgical AI that offers trustworthy guidance to surgeons at every level of experience.
Related
- Traction machine-induced soil compaction: A review of stress transmission, mitigation, and predictive modelling
- When perfect is a warning: explanations expose shortcuts that accuracy hides in cancer gene-expression classifiers
- Agricultural Equipment, Crop Protection, and Pesticide Reduction: Toward a Systemic Transformation of Farming Practices
- Are You Learning Biological Signal or Shortcuts? Auditing and Mitigating Bias in Protein-Protein Interaction Datasets
- Rethinking agricultural equipment design by understanding its uses: three insights for the agroecological transition
BibTeX
@article{protserov2026analysis,
title = {{Analysis and mitigation of equipment-induced shortcuts in AI models for laparoscopic cholecystectomy}},
author = {Sergey Protserov and Anastasiia Repalo and Pouria Mashouri and Jaryd Hunter and Caterina Masino and Amin Madani and Michael Brudno},
journal = {PLOS Digital Health},
year = {2026},
month = oct,
doi = {10.1371/journal.pdig.0001421},
url = {https://journals.plos.org/digitalhealth/article?id=10.1371/journal.pdig.0001421},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Analysis and mitigation of equipment-induced shortcuts in AI models for laparoscopic cholecystectomy
AU - Sergey Protserov
AU - Anastasiia Repalo
AU - Pouria Mashouri
AU - Jaryd Hunter
AU - Caterina Masino
AU - Amin Madani
AU - Michael Brudno
PY - 2026
DA - 2026/10/08
JO - PLOS Digital Health
DO - 10.1371/journal.pdig.0001421
UR - https://journals.plos.org/digitalhealth/article?id=10.1371/journal.pdig.0001421
AB - Machine learning models have seen a lot of success in medical image segmentation domain. However, one of the challenges that they face are confounders or shortcuts: spurious correlations or biases in the training data that affect the resulting models. One example of such confounders for surgical machine learning is the setup of surgical equipment, including tools and lighting. Using the task of identification of safe and dangerous zones of dissection in laparoscopic cholecystectomy images and videos as a use-case, we inspect two equipment-induced biases: the location of surgical tools in the field of view and the direction of lighting, both of which are tightly related to the region of interest. We propose methods for evaluating the severity of these biases based on measurements of model consistency under real or simulated exposure to these shortcuts, and augmentation-based methods for mitigating them. We show that our tool bias mitigations based on pasting tool images in random locations improve the models’ prediction consistency under tool movements by 9 percentage points in the most inconsistent cases, and by 4 percentage points on average. Our lighting bias mitigations based on simulated lighting adjustments help reduce fraction of pixels originally predicted as belonging to the dangerous zone that may flip to safe under light changes from 5% to 1.5%, without compromising segmentation quality. Author summary: Artificial intelligence is increasingly used to support surgeons, for example by highlighting which tissue is safe to cut and which is dangerous during gallbladder removal (laparoscopic cholecystectomy), one of the most common abdominal operations. These systems can, however, learn the wrong lesson. We found that AI models for this task come to rely on where the surgeon places the instruments and where the camera light is pointed, rather than on the anatomy itself. Experienced surgeons tend to position their tools and lighting in predictable places, so the model takes a shortcut. The danger is that a less experienced surgeon, who handles equipment differently, could receive misleading guidance. This flaw is hard to detect because the pattern runs through all of the data, so ordinary testing does not reveal it. We developed a way to measure how strongly a model leans on these shortcuts, and a simple training method that breaks them by inserting instruments and extra lighting at random positions. Our approach makes predictions substantially more consistent without sacrificing accuracy, a step toward surgical AI that offers trustworthy guidance to surgeons at every level of experience.
SN - 2767-3170
ER -