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: Jan 5, 2026

Comparison of Predictive Performance of Three Lymph Node Staging Systems in Colorectal Signet Ring Cell Carcinoma Based on Machine Learning Model
07:13

Comparison of Predictive Performance of Three Lymph Node Staging Systems in Colorectal Signet Ring Cell Carcinoma Based on Machine Learning Model

Published on: April 18, 2025

465

Prediction of Polyp Pathology Using Convolutional Neural Networks Achieves "Resect and Discard" Thresholds.

Robin Zachariah1, Jason Samarasena1,2, Daniel Luba3

  • 1Department of Gastroenterology and Department of Internal Medicine, University of California Irvine Medical Center, Orange, California, USA.

The American Journal of Gastroenterology
|October 26, 2019
PubMed
Summary

Related Concept Videos

You might also read

Related Articles

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

Sort by
Same author

Endoscopic ultrasound-guided transjugular intrahepatic portosystemic shunt? Not exactly.

Gastrointestinal endoscopy·2026
Same author

Short-Term Outcomes of Laparoscopic Hiatal Hernia Repair with Consecutive Transoral Incisionless Fundoplication: A Prospective Multicenter Cohort Study.

Journal of the American College of Surgeons·2026
Same author

Severe Melanosis Coli Associated With Chronic Aloe Vera Supplement Use.

The American journal of gastroenterology·2026
Same author

Prospective cohort study of palliative treatment of esophageal cancer with liquid nitrogen spray cryotherapy.

Gastrointestinal endoscopy·2026
Same author

Outcomes of endoscopic submucosal dissection for superficial esophageal squamous neoplasms: a multicenter North American experience.

Gastrointestinal endoscopy·2025
Same author

International Expert Delphi Consensus on Endoscopic Ultrasound-guided Portosystemic Pressure Gradient: Best Practices and Future Directions.

Clinical gastroenterology and hepatology : the official clinical practice journal of the American Gastroenterological Association·2025

A new AI model accurately diagnoses colorectal polyps in real-time, potentially saving billions. This convolutional neural network (CNN) technology helps determine if polyps need removal or can be left in place, improving patient care.

Area of Science:

  • Gastroenterology
  • Artificial Intelligence
  • Medical Imaging

Background:

  • Accurate in situ diagnosis of diminutive colorectal polyps (≤5 mm) is crucial for cost-effective management.
  • Current diagnostic methods often fail to meet the Preservation and Incorporation of Valuable Endoscopic Innovations (PIVI) initiative standards.
  • Convolutional neural networks (CNNs) show promise for real-time polyp pathology prediction.

Purpose of the Study:

  • To develop and validate a CNN-based optical pathology (OP) model for in situ diagnosis of colorectal polyps.
  • To assess the model's performance against PIVI thresholds for "resect and discard" and "diagnose and leave" strategies.
  • To evaluate the model's accuracy independent of light source (white light or narrow band imaging [NBI]).

Main Methods:

More Related Videos

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images
04:23

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images

Published on: April 21, 2023

2.2K

Related Experiment Videos

Last Updated: Jan 5, 2026

Comparison of Predictive Performance of Three Lymph Node Staging Systems in Colorectal Signet Ring Cell Carcinoma Based on Machine Learning Model
07:13

Comparison of Predictive Performance of Three Lymph Node Staging Systems in Colorectal Signet Ring Cell Carcinoma Based on Machine Learning Model

Published on: April 18, 2025

465
A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images
04:23

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images

Published on: April 21, 2023

2.2K
  • A CNN model was developed using Tensorflow and pretrained on ImageNet, achieving 77 frames per second.
  • A dataset of 6,223 colorectal polyp images underwent 5-fold cross-training and validation.
  • An additional fresh validation set of 634 polyp images was used to assess real-world performance.

Main Results:

  • The OP model achieved a 97% negative predictive value for adenomas in diminutive rectosigmoid polyps.
  • Surveillance interval concordance between the OP model and true pathology was high (93% in original, 94% in fresh validation).
  • Performance was consistent across white light and NBI imaging.

Conclusions:

  • CNN-based OP models are feasible for real-time, in situ diagnosis of colorectal polyps.
  • The developed model surpasses PIVI thresholds for polyp management strategies.
  • Potential benefits include point-of-care adenoma detection and improved surveillance recommendations.