Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Related Concept Videos

Survival Tree

Survival Tree

Survival trees are a non-parametric method used in survival analysis to model the relationship between a set of covariates and the time until an event of interest occurs, often referred to as the "time-to-event" or "survival time." This method is particularly useful when dealing with censored data, where the event has not occurred for some individuals by the end of the study period, or when the exact time of the event is unknown.
Building a Survival Tree
Constructing a...

You might also read

Related Articles

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

Sort by

Same author

Simulation-Free Single-Fraction High-Dose Radiotherapy for Non-Spine Bone and Soft Tissue Metastases.

Practical radiation oncology·2026

Same author

Genetically proxied inhibition of cholesterol-lowering drug targets and survival in HPV-positive and non-HPV driven head and neck cancer: a multicentre MR study.

Cancer epidemiology, biomarkers & prevention : a publication of the American Association for Cancer Research, cosponsored by the American Society of Preventive Oncology·2026

Same author

Auto-segmentation of organs-of-interest clinical acceptability & reproducibility framework in head and neck cancer.

Physics and imaging in radiation oncology·2026

Same author

Heart Health Begins with Community: An Arts-Based Report on the Experiences of Heart Health in the Moosonee Community.

CJC open·2026

Same author

Spread of a heart failure remote patient management program from speciality to community care settings: multiple case study implementation.

BMC health services research·2026

Same author

Sex-Specific Associations between Long-term Air Pollution Exposure and Coronary Atherosclerosis at Cardiac CT.

Radiology·2026

Search research articles

Related Experiment Video

Updated: May 5, 2026

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

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

Published on: October 10, 2018

Shortcut learning in medical AI hinders generalization: method for estimating AI model generalization without

Cathy Ong Ly^1,2,3, Balagopal Unnikrishnan^3,4,5,6, Tony Tadic^7,8

¹Peter Munk Cardiac Centre and Ted Rogers Centre for Heart Research, University Health Network, Toronto, ON, Canada.

NPJ Digital Medicine

|May 14, 2024

Summary

AI models in healthcare often overestimate performance due to data biases. A new method, P_Est, corrects for these biases, improving accuracy estimates for AI generalizability in medical applications.

More Related Videos

Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images

Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images

Published on: October 27, 2023

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

Related Experiment Videos

Last Updated: May 5, 2026

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

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

Published on: October 10, 2018

Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images

Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images

Published on: October 27, 2023

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

Area of Science:

Artificial Intelligence in Medicine
Machine Learning for Healthcare
Biomedical Data Science

Background:

Increasing complexity and volume of healthcare datasets necessitate robust AI models.
Unstructured medical data (images, ECGs, text) are increasingly utilized with advanced AI.
Estimating AI generalizability to new clinical settings without extensive external validation is a significant challenge.

Purpose of the Study:

To address the overestimation of AI model performance in healthcare settings.
To quantify the impact of data acquisition biases (DAB) and shortcut learning on AI generalizability.
To introduce a novel bias-corrected method for estimating external AI accuracy.

Main Methods:

Conducted experiments across 13 diverse healthcare datasets (X-rays, CTs, ECGs, clinical notes, lung sounds).
Investigated shortcut learning caused by hidden data acquisition biases (DAB).
Developed and validated an open-source, bias-corrected external accuracy estimation method (P_Est).

Main Results:

AI model performance was frequently overestimated by an average of 20% due to shortcut learning from DAB.
The proposed P_Est method demonstrated improved external accuracy estimation.
P_Est achieved an average external accuracy estimation within 4% of true accuracy.

Conclusions:

Shortcut learning from data acquisition biases significantly inflates AI performance estimates in healthcare.
The P_Est method provides a more reliable assessment of AI model generalizability.
This bias-corrected estimation is crucial for deploying trustworthy AI in real-world medical applications.