An adaptive QRS detection algorithm for ultra-long-term ECG recordings

Abstract

Background: Accurate detection of QRS complexes during mobile, ultra-long-term ECG monitoring is challenged by instances of high heart rate, dramatic and persistent changes in signal amplitude, and intermittent deformations in signal quality that arise due to subject motion, background noise, and misplacement of the ECG electrodes. Purpose: We propose a revised QRS detection algorithm which addresses the above-mentioned challenges. Methods and Results: Our proposed algorithm is based on a state-of-the-art algorithm after applying two key modifications. The first modification is implementing local estimates for the amplitude of the signal. The second modification is a mechanism by which the algorithm becomes adaptive to changes in heart rate. We validated our proposed algorithm against the state-of-the-art algorithm using short-term ECG recordings from eleven annotated databases available at Physionet, as well as four ultra-long-term (14-day) ECG recordings which were visually annotated at a central ECG core laboratory. On the database of ultra-long-term ECG recordings, our proposed algorithm ADAPTIVE QRS DETECTION 2 showed a sensitivity of 99.90% and a positive predictive value of 99.73%. Meanwhile, the stateof-the-art QRS detection algorithm achieved a sensitivity of 99.30% and a positive predictive value of 99.68% on the same database. The numerical efficiency of our new algorithm was evident, as a 14-day recording sampled at 200 Hz was analyzed in approximately 157 seconds. Conclusions:We developed a new QRS detection algorithm. The efficiency and accuracy of our algorithm makes it a good fit for mobile health applications, ultra-long-term and pathological ECG recordings, and the batch processing of large ECG databases.

QRS detection and classification in Holter ECG data in one inference step

“…We also evaluated three-detectors from the python NeuroKit package—modified Engelse & Zeelenberg 6 , Hamilton detector 4 , and Kalidash detector 7 . Results for these detectors might differ from the performance reported by respective papers since we used all available data from all datasets; we implemented detectors by Elgendi 2 and Malik 3 using respective papers.…”

“…For comparison, we also evaluated used datasets by several publicly available QRS detection methods: by Elgendi 2 , Malik et al 3 , XQRS detector from Python WFDB package 30 , and by Pan and Tompkins 1 . We also evaluated three-detectors from the python NeuroKit package—modified Engelse & Zeelenberg 6 , Hamilton detector 4 , and Kalidash detector 7 .…”

Wave-shape oscillatory model for nonstationary periodic time series analysis

“…1. Apply any suitable beat tracking algorithm (such as [38] if f is an ECG signal) to determine the locations of all oscillatory cycles in x. Suppose there are N resulting cycles.…”

“…Then, we apply a median filter with a window size of 600 ms to the output of the previous median filter. We detect the QRS complexes in the ECG signal by applying a standard high-accuracy QRS detector [38]. Write the pre-processed ECG signal as a vector x ∈ R n , where n = f s × T is the number of samples, f s = 200 Hz is the sampling rate of the signal, and T = 2155 is the duration of the recording in seconds.…”

Detection of ECG Fiducial Points Using Recursive Estimation and Kalman filtering

“…Onset, offset and peak location of ECG waves are known as fiducial points (FPs) [7]. Several algorithms for automatically detecting QRS complexes have been proposed; for instance, using empirical mode decomposi-tion [8], artificial neural networks [9], wavelets [10][11][12][13], reverse biorthogonal wavelet decomposition and nonlinear filtering [14], quadratic filtering [15], locally adaptive weighted total variation denoising [16], regular grammar and deterministic automata [17], combination of interval and trigonometric threshold values [6], and approaches for ultra-long-term ECG recordings [18]. The advantages of Kalman filter have been discussed in several studies regarding QRS complex detection [7,19]; however, the main problem is the initialization of both search locations and operating parameters.…”