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 Videos

Two-Step Ensemble Convolutional Neural Networks for Colonoscopic Biopsy Classification Resembling Pathologists'

Hongjun Yoon1, Mohammad Rizwan Alam2, Nishant Thakur2

  • 1AI Lab, Deepnoid Inc., Seoul, Korea.

Journal of Korean Medical Science
|May 26, 2026
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

Ranking-Aware Multiple Instance Learning for Histopathology Slide Classification: Development and Validation Study.

JMIR medical informatics·2026
Same author

Practice of Cytopathology in Korea: A 40-Year Evolution Through Standardization, Digital Transformation, and Global Partnership.

Diagnostic cytopathology·2025
Same author

Commercially Available Artificial Intelligence Solutions for Gynaecologic Cytology Screening and Their Integration Into Clinical Workflow.

Cytopathology : official journal of the British Society for Clinical Cytology·2025
Same author

Large-Scale Dermatopathology Dataset for Lesion Segmentation: Model Development and Analysis.

Journal of Korean medical science·2025
Same author

Improved Diagnostic Accuracy of Thyroid Fine-Needle Aspiration Cytology with Artificial Intelligence Technology.

Thyroid : official journal of the American Thyroid Association·2024
Same author

Artificial-Intelligence-Assisted Detection of Metastatic Colorectal Cancer Cells in Ascitic Fluid.

Cancers·2024
Same journal

Characteristics of Adolescents and Young Adults Visiting an Emergency Department Due to Self-Harm or Suicide Attempts: Risk Factors for Severe Medical Outcome and Short-Term Revisits.

Journal of Korean medical science·2026
Same journal

Effect of Socioeconomic Status on Lung Function in the Korean General Population.

Journal of Korean medical science·2026
Same journal

Predictors of Treatment Outcomes in Pediatric Graves' Disease.

Journal of Korean medical science·2026
Same journal

Evaluating Large Language Models for Post-Publication Promotion: A Blinded Comparative Study of Social Media Posts in Public Health.

Journal of Korean medical science·2026
Same journal

Values and Preferences of Korean Patients With Facial Palsy: A Cross-Sectional Survey.

Journal of Korean medical science·2026
Same journal

Publication and Retraction Activity in the Field of Extracorporeal Membrane Oxygenation: Origins, Concerns, and Perspectives.

Journal of Korean medical science·2026
See all related articles
This summary is machine-generated.

This study introduces an AI pipeline for classifying colorectal cancer (CRC) subtypes from whole slide images (WSIs). The ensemble deep learning model accurately distinguishes between five diagnostic categories, improving automated CRC diagnosis.

Area of Science:

  • Oncology
  • Computer Science
  • Medical Imaging

Background:

  • Colorectal cancer (CRC) is a major global health concern, with early detection crucial for survival.
  • Artificial intelligence (AI) shows promise in classifying CRC, but previous studies faced limitations with small datasets and non-specific findings.
  • A comprehensive AI pipeline is needed to address challenges in automated CRC diagnosis, including low-quality images.

Purpose of the Study:

  • To develop and validate an ensemble deep learning pipeline for classifying five diagnostic categories of colorectal cancer (CRC).
  • To address limitations of previous AI studies in CRC classification, such as small datasets and lack of specificity.
  • To filter low-quality whole slide images (WSIs) before AI analysis for improved diagnostic accuracy.

Main Methods:

Keywords:
Artificial IntelligenceBiopsyColorectal CancerComputer-Assisted DiagnosisEnsemble Learning

Related Experiment Videos

  • Utilized a dataset of 18,922 CRC WSIs labeled into five categories: non-tumor, hyperplastic polyp, adenoma, adenocarcinoma, and neuroendocrine tumor (NET).
  • Developed a two-stage classification model: a clustering-constrained attention multiple instance learning model followed by EfficientNet.
  • Implemented an ensemble pipeline to mimic pathologist analysis, including a quality control step for WSIs.

Main Results:

  • The ensemble pipeline achieved high performance across multiple CRC classes, with micro-, macro-, and weighted-F1-scores of 86.57%, 83.83%, and 86.86%, respectively.
  • The model demonstrated particular strength in classifying neuroendocrine tumors (NET), achieving an F1-score of 87.18%.
  • The pipeline effectively handled low-quality WSIs, demonstrating robustness for real-world diagnostic applications.

Conclusions:

  • The proposed ensemble deep learning model offers a reliable solution for automated CRC diagnosis, accurately classifying multiple CRC types.
  • This AI approach enhances clinical practice by providing safeguards for exceptional cases and improving diagnostic accuracy.
  • The method is particularly effective for analyzing whole slide images (WSIs) that may be of low quality in clinical settings.