Evaluation of Electrocardiogram Beat Detectors

Atrial fibrillation (AF) increases stroke risk fivefold but can be reduced with anticoagulation if diagnosed. Smartwatches and handheld ECG devices with automated analysis algorithms provide opportunities for AF detection in telehealth settings, but the performance of ECG beat detection algorithms, crucial for identifying AF’s irregular rhythms, remains unclear. This study evaluated 16 open-source algorithms across seven databases, including clinical, telehealth, and synthetic data, assessing performance on sinus rhythm (SR) and AF, high- and low-quality signals, and various noise types. Key findings revealed that unsw (MATLAB) achieved high F1 scores but was slower, while nk (Python) excelled in speed and noisy data. Performance decreased significantly in telehealth versus clinical settings and with lower-quality signals, while hamilt, nk, and unsw performed better on AF than SR. Noise types impacted performance differently, with motion artifacts having the greatest effect, followed by electrode movements and baseline wander. Overall, nk and unsw are recommended for AF screening in telehealth due to consistent performance, and all algorithms, evaluation code, and data are openly available.

QRS detection in single-lead, telehealth electrocardiogram signals: Benchmarking open-source algorithms

The electrocardiogram (ECG) is a vital tool for assessing heart health. Traditionally, ECGs are recorded in clinical settings, but with advances in technology, mobile devices and smartwatches can now be used to record ECGs in daily life. However, ECG recordings from these devices often contain more noise, posing challenges for accurate analysis. In this study, we evaluated 18 different algorithms for detecting heartbeats in ECGs. Our aim was to identify the best-performing algorithm for use with ECGs recorded using mobile devices. We tested each algorithm on 995 ECG recordings and compared their performance against manually-annotated heartbeats. From our analysis, we identified the two best-performing algorithms. These algorithms performed well when analysing high-quality ECGs obtained under clinical supervision and from mobile devices. However, their performance degraded significantly when analysing noisy ECGs from mobile devices. These findings highlight the importance of selecting robust algorithms for ECG analysis, particularly for data collected outside clinical environments. Furthermore, the study demonstrates the need to ensure that only ECGs which can be accurately analysed are used for clinical decision making.

https://journals.plos.org/digitalhealth/article?id=10.1371/journal.pdig.0000538

Read the full paper on PLOS Digital Health