Early Risk Assessment Model for ICA Timing Strategy in Unstable Angina Patients Using Multi-Modal Machine Learning

Aug 5, 2026

Invasive coronary arteriography (ICA) is the gold standard for diagnosing cardiovascular diseases, yet selecting optimal ICA timing in unstable angina (UA) patients remains difficult. UA lacks specific ST-segment or enzyme indicators, so risk must be inferred from heterogeneous, clinically interpretable signals. This study proposes a machine-learning early risk assessment using multi-modal inputs, trained on 640 UA patients, to identify who benefits most from ICA while supporting practical clinical deployment.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages24

Reading Time1 h 0 min

AccessFree to download

Abstract

Background

Objectives

Methods

Results

Conclusions

1. Introduction

Clinical context and diagnostic challenges

ICA and competing invasive strategies

Need for early risk assessment models

arXiv :2408 .04276v1 [ cs .LG] 8 Aug 2024 Early Risk Assessment Model for ICA Timing Strategy in Unstable Angina Patients Using Multi-Modal Machine Learning Candi Zhenga,b,∗, Kun Liuc , Yang Wanga , Shiyi Chenb , Hongli Lic a Department of Mathematics, Hong Kong University of Science and Technology, Clear Water Bay, Hong Kong SAR, China b Department of Mechanics and Aerospace Engineering, Southern University of Science and Technology, Xueyuan Rd 1088, Shenzhen, China c Department of Cardiology, Shanghai General Hospital, Shanghai, China Abstract Background. Invasive coronary arteriography (ICA) is recognized as the gold standard for diagnosing cardiovascular diseases, including unstable angina (UA) . The challenge lies in determining the optimal timing for ICA in UA patients, balancing the need for revascularization in high-risk patients against the potential complications in low-risk ones. Unlike myocardial infarction, UA does not have specific indicators like ST-segment deviation or cardiac enzymes, making risk assessment complex. Objectives. Our study aims to enhance the early risk assessment for UA patients by utilizing machine learning algorithms. These algorithms can potentially identify patients who would benefit most from ICA by analyzing less specific yet related indicators that are challenging for human physicians to interpret. Methods. We collected data from 640 UA patients at Shanghai General Hospital, including medical history and electrocardiograms (ECG) . Machine learning algorithms were trained using multi-modal demographic characteristics including clinical risk factors, symptoms, biomarker levels, and ECG features extracted by pre-trained neural networks. The goal was to stratify ∗ Corresponding author Email address: [drhonglili@126.com](drhonglili@126.com) (Hongli Li) Preprint patients based on their revascularization risk. Additionally, we translated our models into applicable and explainable look-up tables through discretization for practical clinical use. Results. The study achieved an Area Under the Curve (AUC) of 0 .719±0 .065 in risk stratification, significantly surpassing the widely adopted GRACE score’s AUC of 0 .579 ± 0.044. Conclusions. The results suggest that machine learning can provide superior risk stratification for UA patients. This improved stratification could help in balancing the risks, costs, and complications associated with ICA, indicating a potential shift in clinical assessment practices for unstable angina. Keywords: Unstable Angina; Coronary Arteriography; Machine Learning; Risk Stratification; Electrocardiogram 1. Introduction Cardiovascular diseases (CVDs) are the leading cause of death globally [1], with ischemic heart disease (IHD) representing the most prevalent and life-threatening type. IHD typically arises from atherosclerotic narrowing within the coronary arteries, which compromises adequate myocardial perfusion. It primarily manifests as either stable ischemic heart disease (chronic coronary syndrome, CCS) or acute coronary syndrome (ACS) . Among ACS, the Non-ST-elevation acute coronary syndromes (NSTE-ACS), encompassing conditions such as unstable angina (UA) and Non-ST-elevation myocardial infarction (NSTEMI) [2], pose significant diagnostic challenges. This complexity stems from the fact that electrocardiogram (ECG) abnormalities, crucial for ACS diagnosis through the identification of ischemic ST-segment changes, are not always present in NSTE-ACS. The complexity intensifies in the case of UA patients, which spans a wide clinical range from patients with minor arterial narrowing to patients at imminent risk of severe angina or myocardial infarction (MI), necessitating prompt revascularization. Accurately diagnosing UA is essential for determining the optimal timing of such interventions, striving to balance the immediacy of revascularization in high-risk patients against the costs and associated complications. Invasive procedu

FAQs

Why is determining ICA timing in unstable angina challenging?

Unstable angina lacks specific indicators such as ST-segment deviation or cardiac enzymes, making risk assessment difficult and requiring inference from complex, indirect clinical information.

What data and features are used in the proposed model?

The study uses multi-modal inputs from 640 UA patients, including medical history, ECG information, clinical risk factors, symptoms, and biomarker levels, with ECG features extracted by pre-trained neural networks.

How does the proposed model compare with the GRACE risk score?

The model achieves an AUC of 0.719±0.065 for risk stratification, which is higher than the GRACE score’s AUC of 0.579±0.044.

Aug 5, 2026

Upload by Clementine