A machine learning algorithm to predict a culprit lesion after out of hospital cardiac arrest
Nilesh Pareek1,2, Christopher Frohmaier3,4, Mathew Smith4
1King's College Hospital NHS Foundation Trust, London, UK.
Insights
A new machine learning algorithm accurately predicts culprit coronary artery lesions in out-of-hospital cardiac arrest (OHCA) patients. This tool aids in identifying critical blockages, improving patient outcomes and guiding treatment decisions.
Area of Science:
- Cardiology
- Artificial Intelligence
- Medical Informatics
Background:
- Out-of-hospital cardiac arrest (OHCA) poses significant mortality risks.
- Identifying the cause of OHCA, particularly culprit coronary artery lesions, is crucial for timely intervention.
- Current diagnostic methods for OHCA may not always pinpoint the underlying coronary artery disease.
Purpose of the Study:
- To develop and validate a machine learning algorithm for predicting culprit coronary artery lesions in OHCA patients.
- To create a tool that assists clinicians in identifying the cause of cardiac arrest.
- To improve the diagnostic accuracy for coronary artery disease in the context of OHCA.
Main Methods:
- A gradient boosting machine learning model was developed using data from the King's Out-of-Hospital Cardiac Arrest Registry.
- The algorithm was trained on 398 patients and validated on two independent European cohorts (568 patients).
- Key predictors included age, ECG findings, regional wall motion abnormalities, vascular disease history, and initial rhythm.
Main Results:
- The algorithm achieved high accuracy, with an area under the curve (AUC) of 0.89 in the development cohort and 0.83/0.81 in validation cohorts.
- It significantly outperformed electrocardiogram (ECG) alone (AUC: 0.69/0.67/0.67) in predicting culprit lesions.
- The model incorporates nine variables and is presented as a user-friendly web application.
Conclusions:
- A novel, simple machine learning algorithm can accurately predict culprit coronary artery lesions in OHCA patients.
- This algorithm offers a valuable tool for risk stratification and guiding further diagnostic and therapeutic strategies.
- The developed model demonstrates high predictive performance and potential for clinical implementation.
Background:
We aimed to develop a machine learning algorithm to predict the presence of a culprit lesion in patients with out-of-hospital cardiac arrest (OHCA).
Methods:
We used the King's Out-of-Hospital Cardiac Arrest Registry, a retrospective cohort of 398 patients admitted to King's College Hospital between May 2012 and December 2017. The primary outcome was the presence of a culprit coronary artery lesion, for which a gradient boosting model was optimized to predict. The algorithm was then validated in two independent European cohorts comprising 568 patients.
Results:
A culprit lesion was observed in 209/309 (67.4%) patients receiving early coronary angiography in the development, and 199/293 (67.9%) in the Ljubljana and 102/132 (61.1%) in the Bristol validation cohorts, respectively. The algorithm, which is presented as a web application, incorporates nine variables including age, a localizing feature on electrocardiogram (ECG) (≥2 mm of ST change in contiguous leads), regional wall motion abnormality, history of vascular disease and initial shockable rhythm. This model had an area under the curve (AUC) of 0.89 in the development and 0.83/0.81 in the validation cohorts with good calibration and outperforms the current gold standard-ECG alone (AUC: 0.69/0.67/0/67).
Conclusions:
A novel simple machine learning-derived algorithm can be applied to patients with OHCA, to predict a culprit coronary artery disease lesion with high accuracy.
Related Concept Videos
01:25Cardiopulmonary Resuscitation IV: Pharmacological Management
01:23Cardiopulmonary Resuscitation III: AED Use
01:30Acute Coronary Syndrome III: Diagnostic Studies