Key result
The ENCASE method combining deep neural networks and engineered features achieved F1 scores of 0.9117 for Normal, 0.8128 for Atrial Fibrillation, and 0.7505 for Others on the hidden test set.
Why the study?
The study aimed to combine deep neural networks and domain-knowledge-based engineered features for cardiac arrhythmia detection from short single-lead ECG recordings.
Combining deep neural networks with engineered features yields high performance for automated cardiac arrhythmia detection from single-lead ECGs.
No takes yet. Share an insight, caveat, or question.
May facilitate automated single-lead ECG screening; leaves open prospective clinical validation.
Hong et al. (2019) studied cardiac arrhythmia. ENCASE (deep neural networks and engineered features) vs. individual features was evaluated on F1 score on hidden test set. The ENCASE method combining deep neural networks and engineered features achieved F1 scores of 0.9117 for Normal, 0.8128 for Atrial Fibrillation, and 0.7505 for Others on the hidden test set.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: