.png){kind=icon}
.png){kind=icon}
Return
Data-driven pediatric ECG reference intervals with VSD-based validation
.png){kind=icon}
PRE
AI
DOI：10.1088/1361-6579/ae3c56.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Objective. To establish population-specific, age- and sex-stratified electrocardiographic (ECG) reference ranges for Chinese children and adolescents using a data-driven approach, addressing the limitations of conventional empirically defined age groupings. Approach. A total of 35 088 ECG recordings from individuals under 18 years of age without structural heart disease or ECG abnormalities were analyzed. An unsupervised machine-learning clustering algorithm was applied to identify natural developmental trajectories of 149 ECG parameters and derive data-driven age intervals. Sex-specific stratification was performed to account for physiological differences. To assess physiological validity, we evaluated the ability of the newly derived reference ranges to identify ECG deviations in children with echocardiographically confirmed ventricular septal defects (VSDs). Main Results. Four distinct age-dependent variation patterns were identified across the 149 ECG parameters, enabling precise determination of age-specific intervals. Sex-related differences were observed for most measurements. When applied to children with VSD, the data-driven reference intervals demonstrated higher sensitivity in detecting ECG deviations compared with previously published standards. Significance. This study introduces a machine-learning-based paradigm for defining pediatric ECG reference values. The resulting age- and sex-specific thresholds more accurately reflect physiological maturation and cardiac loading changes than traditional reference sets, offering improved clinical relevance for pediatric ECG interpretation.
Keywords:
electrocardiogram
reference range
age-specific
data-driven
clustering
Journal
P
IF:
2.7
Papers:
103
Citations:
7.3K