Key result
Machine learning outperforms classical ECG criteria in detecting LVH in young males, achieving AUC ~0.87.
Why the study?
Classical ECG criteria for LVH are well studied in older and hypertensive populations, but their utility in young pre-participation cohorts remains unclear.
Do machine learning models improve the detection of echocardiographic left ventricular hypertrophy from ECG compared to classical criteria in young males undergoing pre-participation screening?
Population
17 310 males aged 16 to 23 screened prior to military conscription
Comparison
Machine learning models vs classical ECG criteria for LVH detection
Design
Observational diagnostic study
Authors
Loading...
May support ML-enhanced ECG screening for LVH in young males; hypothesis-generating and requires prospective validation before adoption.
Observational (n=17,310)
Do machine learning models improve the detection of echocardiographic left ventricular hypertrophy from ECG compared to classical criteria in young males undergoing pre-participation screening?
Effect estimate: AUC 0.873 (95% CI 0.817-0.929)
Machine learning models applied to screening ECGs significantly outperform classical ECG criteria for detecting echocardiographic left ventricular hypertrophy in young males.
Lim et al. (2021) conducted an observational in Left ventricular hypertrophy (n=17,310). Machine learning models vs. Classical ECG criteria was evaluated on Detection of echocardiogram-diagnosed left ventricular hypertrophy (AUC 0.873, 95% CI 0.817-0.929). Machine learning models, such as GLMNet (AUC 0.873; 95% CI 0.817-0.929), were superior to classical ECG criteria for detecting echocardiographic left ventricular hypertrophy in young males.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: