Key result
C5.0 algorithm achieves ~95% balanced accuracy classifying cardiac rhythms using interpretable ECG features.
Why the study?
Analysis of heart rate variability in patients with inappropriate sinus tachycardia was performed to understand changes in temporal and spectral indexes and underlying mechanisms.
Observational (n=10,646)
Yes
May support interpretable automated ECG classification; leaves open prospective validation before clinical use.
Analysis of heart rate variability in patients with inappropriate sinus tachycardia showed a 24-hour decrease in all temporal and spectral indexes, even after attempted correction to a rate of 75 beats/min. This may have resulted from a global decrease in parasympathetic activity or from a rapid sinus rate produced by other ill-defined mechanisms.
No takes yet. Share an insight, caveat, or question.
Castellanos et al. (1998) conducted an observational in Cardiac arrhythmias (n=10,646). C5.0 decision tree algorithm vs. Other machine learning models (GLM, Logit, k-NN, Naive Bayes, Random Forest, XGBoost) was evaluated on Balanced accuracy for classifying four classes of cardiac rhythms. The C5.0 white-box machine learning algorithm achieved a balanced accuracy of 95.35% in classifying four classes of cardiac rhythms using only five interpretable ECG features.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: