Annu Int Conf IEEE Eng Med Biol Soc. 2022 Jul;2022:2013-2016. doi: 10.1109/EMBC48229.2022.9871435.
An algorithm to detect P- and T-waves in an electrocardiogram (ECG) signal is presented. The algorithm has physical origins inspired by weak signal detection by leveraging stochastic resonance (SR) in a well potential. Specifically, a particle inside an underdamped monostable well is introduced with the ECG signal. The parameters defining the well and system characteristics are optimized towards enhancing the P-, R-, and T -waves while suppressing the other portions including noise-only sections. The enhanced features are detected by thresholding. Based on the performance obtained from the QT database, the algorithm achieves an average sensitivity of 99.97% for P-waves and an average sensitivity of 99.35% for T-waves, better than most P- and T-wave detection algorithms reported. Clinical Relevance- The proposed SR algorithm achieves high P- and T-wave detection performance and can potentially be integrated with implantable long-term cardiac monitors for patients experiencing rare symptoms without deteriorating the battery life.
提出了一种用于检测心电图（ECG）信号中 P 波和 T 波的算法。该算法具有物理起源，灵感来自于在势阱中利用随机共振（SR）进行弱信号检测。具体来说，将心电图信号引入到欠阻尼单稳阱中的一个粒子中。定义阱和系统特性的参数被优化，以增强 P 波、R 波和 T 波，同时抑制包括仅噪声部分在内的其他部分。通过阈值检测增强的特征。基于从 QT 数据库获得的性能，该算法对 P 波的平均灵敏度为 99.97%，对 T 波的平均灵敏度为 99.35%，优于大多数报告的 P 波和 T 波检测算法。临床意义——所提出的 SR 算法实现了高的 P 波和 T 波检测性能，并且可以与植入式长期心脏监测器集成，用于经历罕见症状的患者，而不会降低电池寿命。
