Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
April 10, 2026BMJ OpenOpen Access

Machine-learning models will be developed and validated to differentiate angina with no obstructive coronary artery disease from obstructive disease and predict 1-, 3-, and 5-year mortality.

View Full Paper
Ask AI
Bookmark
Share

Population

Adults undergoing initial cardiac catheterisation for chest pain/anginal equivalents since 1995, excluding…

Design

Other

Follow-up

1, 3 and 5 years

Key result

Machine-learning models will be developed and validated to differentiate angina with no obstructive coronary artery disease from obstructive disease and predict 1-, 3-, and 5-year mortality.

Authors

JDJiawen DengSPShubh PatelMFMarinda Fung

Discussion

Loading...

Member takes

Overview

This study protocol describes the planned development and validation of machine-learning models to diagnose ANOCA and predict long-term mortality in patients undergoing cardiac catheterization.

Key Points

  • The aim is to develop machine-learning models that distinguish between ANOCA and obstructive CAD and predict long-term mortality for both conditions.
  • Develop classification models using a multicentre retrospective cohort.
  • Eligibility includes adults with chest pain undergoing cardiac catheterisation.
  • Use nested cross-validation and site-specific methods for model validation.
  • Preprocess data with missing-data imputation and feature scaling.
  • Employ various algorithms and optimise hyperparameters.
  • Models differentiate ANOCA from obstructive CAD based on anatomical stenosis severity.
  • Mortality prediction models aim to establish 1, 3, and 5-year mortality rates for ANOCA and obstructive CAD.
  • Final models will serve as a web-based clinical risk calculator.

Structured PICO

P
Population
Adults (≥18 years) undergoing initial cardiac catheterisation for chest pain/anginal equivalents since 1995, excluding prior revascularisation, major structural heart disease and predefined non-anginal indications.
I
Intervention
Machine-learning classification models (elastic-net logistic regression, random forest, LightGBM and multilayer perceptron models)
O
Outcome
Diagnosis of ANOCA (0% to <50% stenosis) versus obstructive CAD (≥50% stenosis) and 1, 3 and 5-year mortality

This study protocol describes the planned development and validation of machine-learning models to diagnose ANOCA and predict long-term mortality in patients undergoing cardiac catheterization.

Cite This Study

Deng et al. (2026) studied this question. Machine-learning models will be developed and validated to differentiate angina with no obstructive coronary artery disease from obstructive disease and predict 1-, 3-, and 5-year mortality.

synapsesocial.com/papers/69d895796c1944d70ce067adhttps://doi.org/10.1136/bmjopen-2025-108799
View Full Paper
Ask AI
Bookmark
Share