An Adaptive and Time-Efficient ECG R-Peak Detection Algorithm

Q: What is the clinical evidence from this study?

Population: Cardiac arrhythmia and real-world ECG morphology variability (n=130). Intervention: Adaptive R-peak detection algorithm using wavelet multiresolution analysis (WMRA), signal mirroring, and adaptive amplitude/time interval thresholding vs. Traditional Pan-Tompkins R-peak detection method. Primary outcome: R-peak detection accuracy and mean processing time per ECG record.

Q: What does this research mean for the field?

An adaptive wavelet multiresolution analysis-based R-peak detection algorithm achieves over 98.8% accuracy on standard benchmarks while reducing processing time by over 30% compared to the traditional Pan-Tompkins method. Novelty: ClaimNovelty.METHODOLOGICAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

Qin Qin; Jianqing Li; Yinggao Yue

doi:10.1155/2017/5980541

January 1, 2017Journal of Healthcare EngineeringOpen Access

An Adaptive and Time-Efficient ECG R-Peak Detection Algorithm

View Full Paper

Ask AI

Bookmark

Share

Key result

Adaptive WMRA algorithm speeds R-peak detection by ~32% vs. Pan-Tompkins with >98% accuracy.

n=130

Figures

Why the study?

Does an adaptive wavelet multiresolution analysis and thresholding algorithm improve R-peak detection accuracy and reduce computation time compared to the traditional Pan-Tompkins method in ECG records from the MIT-BIH arrhythmia and QT databases?

Key Points

Develop an adaptive and computationally efficient algorithm to accurately detect R-peaks in electrocardiogram signals without requiring complex search-back routines or prelearning procedures.
Applied wavelet multiresolution analysis to enhance ECG signals, mirrored signals to flip negative R-peaks, and identified local maximums using a first-order forward differential approach.
Applied adaptive amplitude and time interval thresholds with initial coefficients K amp of 0.25 and K time of 0.45 across a 10 s shifting window.
Evaluated detection accuracy and processing speed against the Pan-Tompkins method using the MIT-BIH arrhythmia database (109966 beats across 48 records) and the QT database (86995 beats).
Achieved average sensitivity of 99.39%, positive predictivity of 99.49%, and accuracy of 98.89% on the MIT-BIH arrhythmia database, with 1229 total detection failures (668 false negatives and 561 false positives).
Achieved average sensitivity of 99.83%, positive predictivity of 99.90%, and accuracy of 99.73% on the QT database, with 238 total detection failures (147 false negatives and 91 false positives).
Reduced processing time to 0.872 s per record on the MIT-BIH database and 0.763 s on the QT database, achieving 30.6% and 32.9% time reductions compared to the Pan-Tompkins method (1.256 s and 1.137 s, respectively).

Structured PICO

Does an adaptive wavelet multiresolution analysis and thresholding algorithm improve R-peak detection accuracy and reduce computation time compared to the traditional Pan-Tompkins method in ECG records from the MIT-BIH arrhythmia and QT databases?

P

Population

130 ECG records comprising 196,961 beats from the MIT-BIH arrhythmia database (48 30-minute records) and the QT database (82 15-minute records) with varied QRS and ST-T morphologies and noise.

I

Intervention

Adaptive and time-efficient ECG R-peak detection algorithm combining 8-level bior6.8 wavelet multiresolution analysis (WMRA), soft thresholding, negative R-peak signal mirroring, first-order forward differential local maximum detection, and adaptive amplitude and time interval thresholding (Kamp=0.25, Ktime=0.45, 10 s shifting window).

C

Comparator

Traditional Pan-Tompkins R-peak detection method and other published R-peak detection algorithms.

O

Outcome

R-peak detection performance (sensitivity [SEN], positive predictivity [+P], and accuracy [ACC] within a 50 ms tolerance window) and mean processing time consumption per ECG record.surrogate

Limitations

Higher or equal time consumption compared to the Pan-Tompkins method in records with numerous large T or P waves or large negative amplitudes (e.g., MITDB records 107, 109, 113, and 116)
High number of detection failures (314 false negatives and 328 false positives) in MITDB record 207 due to severely distorted heartbeats
Amplitude threshold may fail to detect small R-peaks when mixed in segments containing numerous large R-peaks (e.g., MITDB records 105, 106, 108, and 228)
Only the first ECG channel from both databases was used for algorithm development and testing
Higher time complexity and processing time in ECG records with numerous large T waves, P waves, or large negative amplitudes requiring frequent time interval comparisons (records 107, 109, 113, and 116 in MITDB)
Substantial detection errors on MITDB record 207 (314 false negatives and 328 false positives) due to severely distorted heartbeats
Methodological defect where the adaptive amplitude threshold fails to detect small R-peaks mixed within segments containing numerous large R-peaks (records 105, 106, 108, and 228)
Evaluation was restricted to only the first channel of the two-channel ECG recordings in both databases

Ask about this paper

Answers are grounded in this paper and cite their sources.

Implication

This computational study demonstrates that combining wavelet multiresolution analysis with signal mirroring and adaptive thresholding achieves >98.8% R-peak detection accuracy while reducing processing time by over 30% compared to the Pan-Tompkins algorithm, supporting real-time ambulatory ECG monitoring.

Abstract

R-peak detection is crucial in electrocardiogram (ECG) signal analysis. This study proposed an adaptive and time-efficient R-peak detection algorithm for ECG processing. First, wavelet multiresolution analysis was applied to enhance the ECG signal representation. Then, ECG was mirrored to convert large negative R-peaks to positive ones. After that, local maximums were calculated by the first-order forward differential approach and were truncated by the amplitude and time interval thresholds to locate the R-peaks. The algorithm performances, including detection accuracy and time consumption, were tested on the MIT-BIH arrhythmia database and the QT database. Experimental results showed that the proposed algorithm achieved mean sensitivity of 99.39%, positive predictivity of 99.49%, and accuracy of 98.89% on the MIT-BIH arrhythmia database and 99.83%, 99.90%, and 99.73%, respectively, on the QT database. By processing one ECG record, the mean time consumptions were 0.872 s and 0.763 s for the MIT-BIH arrhythmia database and QT database, respectively, yielding 30.6% and 32.9% of time reduction compared to the traditional Pan-Tompkins method.

Discussion

No takes yet. Share an insight, caveat, or question.

Cite This Study

Qin et al. (2017) studied Cardiac arrhythmia and real-world ECG morphology variability (n=130). Adaptive R-peak detection algorithm using wavelet multiresolution analysis (WMRA), signal mirroring, and adaptive amplitude/time interval thresholding vs. Traditional Pan-Tompkins R-peak detection method was evaluated on R-peak detection accuracy and mean processing time per ECG record. The adaptive WMRA-based R-peak detection algorithm achieved 98.89% accuracy on MITDB and 99.73% on QTDB while reducing processing time by 30.6% and 32.9% versus the Pan-Tompkins method.

synapsesocial.com/papers/6a8dfd548dae308ac772cd63 https://doi.org/10.1155/2017/5980541

View Full Paper

Ask AI

Bookmark

Share

An Adaptive and Time-Efficient ECG R-Peak Detection Algorithm

Figures

Key Points

Structured PICO

Limitations

Ask about this paper

Implication

Abstract

Discussion

Cite This Study

An Adaptive and Time-Efficient ECG R-Peak Detection Algorithm

Figures

Key Points

Structured PICO

Limitations

Ask about this paper

Implication

Abstract

Discussion

Cite This Study

Keep reading

Keep reading