Key result
FINESSE protocol aims to develop machine learning models predicting mortality and major cardiovascular events after stress echocardiography.
Why the study?
Coronary artery disease remains a leading cause of morbidity and mortality, necessitating accurate diagnostic strategies and robust risk stratification in patients undergoing stress echocardiography.
Does a machine learning-based model improve risk prediction compared to standard risk scores in patients undergoing stress echocardiography for suspected coronary artery disease?
Observational
Yes
Does a machine learning-based model improve risk prediction compared to standard risk scores in patients undergoing stress echocardiography for suspected coronary artery disease?
The FINESSE study protocol outlines the development and validation of machine learning models to enhance risk stratification in patients undergoing stress echocardiography for suspected coronary artery disease.
Answers are grounded in this paper and cite their sources.
Protocol for ML-based risk models in stress echocardiography; leaves open whether they improve prediction over standard scores pending validation.
Aims: Coronary artery disease (CAD) remains a leading cause of morbidity and mortality worldwide, necessitating accurate diagnostic strategies and robust risk stratification. The FINESSE (Artificial Intelligence Stress Echo) study aims to develop and validate machine learning-based models to improve risk prediction in patients undergoing stress echocardiography (SE) for the assessment of inducible myocardial ischaemia, and to implement a structured AI-driven risk reclassification framework to support clinical decision-making. Methods and analysis: This is a retrospective observational study on prospectively recruited patients referred for SE with suspected CAD. Clinical, demographic, haemodynamic, and echocardiographic data will be extracted from institutional electronic health records and linked with longitudinal national datasets. The study employs a three-stage ML framework: (1) baseline risk estimation using QRISK3, (2) development and optimization of supervised ML models (including ensemble methods, support vector machines, and neural networks), and (3) longitudinal risk prediction using linked NHS England data (such as all-cause mortality and major adverse cardiovascular events, including cardiovascular death, non-fatal myocardial infarction, stroke, unplanned revascularization). Model performance will be evaluated using discrimination (AUC), calibration, and reclassification metrics, and will be externally validated using independent multicentre SE datasets. Conclusion: The FINESSE study will evaluate the incremental value of machine learning in SE by integrating multimodal data into a clinically actionable risk reclassification framework. This approach has the potential to improve personalized risk stratification, optimize diagnostic pathways, and support precision cardiovascular care. Protocol registration: Ethical approval has been obtained from the UK Health Research Authority and relevant research ethics committees, with appropriate data governance approvals for the use of routinely collected healthcare data. The FINESSE study is registered on ClinicalTrials.gov (Identifier: NCT07432620), ensuring transparency and public accessibility of the study protocol.
No takes yet. Share an insight, caveat, or question.
Ihekwaba et al. (2026) conducted an observational in Suspected coronary artery disease. Machine learning-based risk prediction models vs. QRISK3 was evaluated on All-cause mortality and major adverse cardiovascular events (cardiovascular death, non-fatal myocardial infarction, stroke, unplanned revascularization). The FINESSE study is a protocol to develop and validate machine learning-based models for predicting mortality and major adverse cardiovascular events in patients undergoing stress echocardiography.
Synapse has enriched 10 related papers. Consider them for comparative context: