Collection of pediatric ECG data for testing detection algorithms in Automated External Defibrillators

Arrhythmia detection algorithms in Automated External Defibrillators (AEDs) need a special approval for use in children aged 0-8 years. Our aim is to establish a pediatric ECG reference dataset with rhythm annotations for the assessment of arrhythmia detection algorithms of AEDs. The database will consist of a training dataset with a public interface for end-users to optimize their algorithms and an independent validation data set that remains hidden. Currently we collected and analyzed 534 pediatric ECGs with non-shockable heart rhythms. We characterized the signal automatically by estimating noise level, power line interferences and movement artifacts. Also RR intervals and extrasystoles are automatically annotated. In combination with additional clinical annotations the pediatric ECG dataset provides an instrument for the development and assessment of AED algorithms for arrhythmia detection in children.