一种可在胸外按压期间进行室颤节律辨识的新算法

发布时间：2018-07-17 15:01

【摘要】：在心肺复苏期间,由于胸外按压对心电(ECG)信号产生了机械干扰,无法可靠地辨识心电节律。而中断胸外按压会减小复苏成功的可能性,所以本文研究开发了一个新的滤波算法——增强最小均方(eLMS)算法,可以在无需硬件参考信号支持的情况下,成功滤除胸外按压干扰,达到在不间断胸外按压的情况下正确辨识室颤(VF)节律和正常窦性(SR)节律。该滤波算法仅用受按压干扰的心电(cECG)信号实现滤波,无需其它参考信号。通过在不同信噪比的情况下混合ECG信号和按压干扰信号来验证该算法,并且与其它已经成熟的算法比较。验证结果表明在不同信噪比情况下,eLMS方法的辨识结果均优于其他方法。进一步研究表明,仅用cECG信号就可以很好地辨识心电节律。本算法的成功研发降低了体外除颤仪的研发成本,提高了心电节律辨识的准确性以及复苏成功的可能性。
[Abstract]:During cardiopulmonary resuscitation (CPR), electrocardiogram (ECG) rhythm can not be recognized reliably due to mechanical interference of external chest compression on ECG signal. The possibility of successful resuscitation will be reduced by interruption of chest compression. Therefore, a new filtering algorithm, enhanced least mean Square (eLMS) algorithm, is developed in this paper, which can successfully filter out the extrathoracic compression interference without the support of the hardware reference signal. The rhythm of ventricular fibrillation (VF) and normal sinus (SR) can be correctly identified under uninterrupted external pressure. The filtering algorithm only uses the ECG signal which is interfered with by compression, and no other reference signal is needed. The algorithm is verified by mixing ECG signals and pressing interference signals under different SNR, and compared with other mature algorithms. The results show that the identification results of ELMS are superior to those of other methods under different signal-to-noise ratio (SNR). Further studies show that ECG signals can be used to identify ECG rhythm well. The successful development of this algorithm reduces the development cost of external defibrillator, improves the accuracy of ECG rhythm identification and the possibility of successful resuscitation.
【作者单位】：军事医学科学院卫生装备研究所;
【基金】：国家自然科学基金资助项目(81501551) 国家科技支撑计划资助项目(2012BAI20B01)
【分类号】：R459.7;TN911.7

【相似文献】