A Detector for Premature Atrial and Ventricular Complexes

Abstract

The relationship between premature atrial complexes (PACs) and atrial fibrillation (AF), stroke and myocardium degradation is unclear. Current PAC detectors are beat classifiers that attain low sensitivity on PAC detection. The lack of a proper PAC detector hinders the study of the implications of this event and its monitoring. In this work a PAC and ventricular detector is presented. Two PhysioNet open-source databases were used: the long-term ST database (LTSTDB) and the supraventricular arrhythmia database (SVDB). A combination of heart rate variability (HRV) and morphological features were used to classify beats. Morphological features were extracted from the ECG as well as on the 4th scale of the discrete wavelet transform (DWT). After feature selection, a random forest algorithm was trained for a binary classification of PAC (S) vs. others and for a multi-labels classification to discriminate between normal (N), S and ventricular (V) beats. The algorithm was tested in a 10-fold cross-validation following a patient-wise train-test division (i.e., no beats belonging to the same patient were included both in the test and train set). The resultant median sensitivity, specificity and positive predictive value (PPV) were 99.29, 99.54, and 100% for (N), 95.83, 99.39, and 35.68% for (S), 100, 99.90, and 79.63% for (V). The proposed method attains a greater PAC and ventricular beat sensitivity and PPV than the state-of-the-art classifiers.

Detection of Non-Sustained Supraventricular Tachycardia in Atrial Fibrillation Screening

“…However, the majority of detectors for supraventricular/premature atrial beats are, in fact, beat classifiers known to provide low sensitivity [12] , [15] . The detector proposed for premature atrial beats in [13] , which yielded a high sensitivity in SVADB, uses two-lead ECG signals. In the present study, the proposed method solves both quality control and ectopic beat handling in one single step, achieved by a design which is independent of beat morphology and lead selection.…”

“…Detection of supraventricular arrhythmia has often been addressed by employing feature-based classification. The features are exemplified by RR intervals in successive short windows, wave amplitudes, wave durations, crosscorrelation to a template beat [10] , [11] , [12] , [13] , and spatial features derived from the vectorcardiogram [14] . Classification has been performed using random forests, support vector machines, linear discriminants, and neural networks [15] .…”

Artificial intelligence for atrial fibrillation detection, prediction, and treatment: A systematic review of the last decade (2013–2023)

“…Bashar et al (2020) and Bashar, Han, et al (2021) combined features from P-wave characteristics and ECG R-R interval variability to detect AF. García-Isla et al (2021) and Gupta et al (2023), employed local mean decomposition to decompose ECG data and extract entropy-based features, then they used an ensemble boosted trees classifier for AF detection. Hartikainen et al (2019) used ECG data from Holter recordings using entropy measures to detect AF.…”

“…. (2020); Buscema et al (2020); Domazetoski et al (2022); He et al (2022); Huang et al (2021); Nickelsen et al (2017); Pérez-Valero et al (2019); Rad et al (2021); Rouhi et al (2021); Schaefer et al (2014); Wesselius et al (2022); Yao et al (2021), Abdul-Kadir et al (2016),Asgari et al (2015),Daqrouq et al (2014),Fuadah and Lim (2022),Bashar et al (2020), Bashar, Han, et al, 2021, García-Isla et al (2021,Gupta et al (2023),Stergiou et al (2018),Li, Feng, et al (2019),Kalidas and Tamil (2019),Keidar et al (2021),Lee et al (2013), Li et al (2021), Mateo and Joaquín Rieta (2013), Dörr et al (2019), Brüser et al (2013), Zalabarria et al (2020), Baalman et al (2020), Bu s et al (2022), Lown et al (2020), Duan et al (2022), Luongo et al (2022), Eerikainen et al (2020), Liao et al (2022), Lahdenoja et al (2018), Navoret et al (2013), Piekarski et al (2016), Mehrang et al (2019) DL 56 37% Biton et al (2023); Christopoulos et al (2022); Fern andez-Ruiz (2019); Lai et al (2020); Ma and Xia (2023); Mittal et al (2021); Mousavi et al (2021); Orchard et al (2016); Rahul and Sharma (2022); Seo et al (2021); M. U. Yang et al (2022), Pourbabaee et al (2018). H. Wu et al (2022), X. Zhang, Jiang, et al (2023), Prabhakararao and Dandapat (2022), Gahungu et al (2021), X. Chen et al (2021), M. Jiang et al (2021); Z. Li, Feng, et al (2019); Yu et al (2022); O. Zhang, Ding, et al (2021), Fang et al (2023), Ukil et al (2022), Kumar et al (2022), Kong et al (2019), Baek et al (2021); Cai et al (2020), Cao et al (2019), Ribeiro et al (2020), Petroni et al (2023), P. Zhang et al (2022); X. Zhang, Jiang, et al (2023), Fiorina et al (2022), Ríos-Muñoz et al (2022), Diamant et al (2022), Shi et al (2020), Weimann and Conrad (2021), Xiong et al (2022), B. Chen et al (2023), Asadi et al (2023), Jo et al (2021), Mousavi et al (2020), Salinas-Martínez et al (2021), B. Chen et al (2022), Ivaturi et al (2021), Parmar et al (2023), C amara-V azquez et al (2021), S. Liu, Wang, et al (2022), Kwon et al (2020), Kudo et al (2023), Kwon et al (2020), Cheng et al (2020), O. Zhang, Ding, et al (2021), Wasserlauf e...…”