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 Concept Videos

You might also read

Related Articles

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

Sort by
Same author

Visible-light-enabled three-component carboamidation of alkenes with aryl thianthrenium salts.

Chemical science·2026
Same author

Machine learning-based integration develops a novel lysosome-related prognostic signature associated with prognosis and immune infiltration landscape in acute myeloid leukemia.

Discover oncology·2026
Same author

Identifying Negative Biomarkers of CD4 Cytotoxic T Lymphocytes via Granzyme B Exclusion Analysis.

European journal of immunology·2026
Same author

Accessing Versatile 4-Aminoprop-3-enylidene Indolones via a 1,6-Nucleophilic Addition/6π-Electrocyclic-Ring-Opening Cascade.

The Journal of organic chemistry·2026
Same author

Correction: Lipid Metabolomics Analysis of a Rosacea Animal Model Induced by Demodex Mite Extract.

Acta parasitologica·2026
Same author

Factors associated with in-hospital mortality in acute exacerbations of COPD: a logistic regression and nomogram model study.

Frontiers in medicine·2026

Related Experiment Video

Updated: Jul 16, 2026

Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma
04:09

Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma

Published on: October 10, 2018

Toward Interpretable Machine Learning Models for Grading Prostate Cancer: A Multicenter Development and Validation

Siyu Dai1, Dewei Lu2, Wei Zheng3

  • 1Graduate School, Guilin Medical University, Guilin, China; Department of Radiology, Nanxishan Hospital of Guangxi Zhuang Autonomous Region, Guilin, China.

Clinical Genitourinary Cancer
|July 4, 2026
PubMed
Summary

An interpretable machine learning model accurately predicts prostate cancer (PCa) grading by integrating radiomic features. This tool aids in distinguishing between lower-risk (Gleason Score ≤ 3+4) and higher-risk (Gleason Score ≥ 4+3) prostate cancer.

Keywords:
InterpretabilityMediation analysisPeritumoral regionRadiomicsSHAP analysis

More Related Videos

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

Introduction of an Integrated Pathology Image Management, Artificial Intelligence, and Reporting System
05:33

Introduction of an Integrated Pathology Image Management, Artificial Intelligence, and Reporting System

Published on: July 11, 2025

Related Experiment Videos

Last Updated: Jul 16, 2026

Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma
04:09

Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma

Published on: October 10, 2018

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

Introduction of an Integrated Pathology Image Management, Artificial Intelligence, and Reporting System
05:33

Introduction of an Integrated Pathology Image Management, Artificial Intelligence, and Reporting System

Published on: July 11, 2025

Area of Science:

  • Radiology and Medical Imaging
  • Machine Learning in Oncology
  • Biomedical Data Science

Background:

  • Accurate grading of prostate cancer (PCa) is crucial for treatment decisions.
  • Distinguishing between Gleason Score (GS) ≤ 3+4 and GS ≥ 4+3 is clinically significant.
  • Radiomic features offer potential for non-invasive cancer characterization.

Purpose of the Study:

  • To develop and validate an interpretable machine learning (ML) model for PCa grading.
  • To differentiate PCa with GS ≤ 3+4 from GS ≥ 4+3.
  • To explore the predictive value and biological significance of radiomic features.

Main Methods:

  • Retrospective multicenter cohort of 225 PCa patients.
  • Manual segmentation of intratumoral and peritumoral regions on T2-weighted imaging.
  • Development of ML models integrating clinical and radiomic features, with interpretability via SHAP analysis.

Main Results:

  • The combined model using clinical and radiomic features achieved high performance (AUC up to 0.975 training, 0.795 external validation).
  • SHAP analysis identified a core predictive radiomic feature.
  • This feature correlated with high GS and partially mediated the PI-RADS score-GS relationship.

Conclusions:

  • An interpretable ML model shows excellent performance for PCa grading.
  • Radiomic features hold significant predictive value for prostate cancer aggressiveness.
  • This approach can aid in non-invasive PCa risk stratification.