Predicting the success of defibrillation by electrocardiographic analysis
Predicting the success of defibrillation by electrocardiographic analysis
复制标题
DOI：
10.1016/s0300-9572(01)00488-9
复制
发表时间：
2002-04-01
期刊：
影响因子：
6.5
通讯作者：
Barbatsis, A
中科院分区：
文献类型：
--
作者：
Povoas, HP;Weil, MH;Barbatsis, A
背景：我们研究了一种心电图信号分析技术，用于预测电击是否会逆转心室颤动（VF），以尽量减少心肺复苏术中重复电击的破坏性影响。方法与结果：采用已建立的心肺复苏模型。采用电致麻醉40 kg家猪VF。在未经治疗的室性心动过速4或7分钟后尝试除颤。由于未能逆转室颤，在随后的每次除颤尝试前，心前压迫和机械通气间隔1分钟。用傅立叶分析计算了数字滤波后的VF小波在不间断心前压缩时的幅频频谱，这些压缩来自传统的右锁骨下电极和左根尖电极。在总共34次电除颤尝试中，24只动物恢复了自发循环（ROSC）。振幅谱分析（AMSA）值为21 mV Hz时，其负预测值为0.96，正预测值为0.78。结论：AMSA预测心肺复苏术中电击未能恢复自发循环时具有较高的阴性预测值。这种方法潜在地满足了最小化无效除颤尝试及其对心肌的不良影响的需要。(C) 2002爱思唯尔科学爱尔兰有限公司版权所有。
Background: We investigated an electrocardiographic signal analysis technique for predicting whether an electrical shock would reverse ventricular fibrillation (VF) in an effort to minimize the damaging effects of repetitive shocks during CPR. Methods and results: An established model of CPR was utilized. VF was electrically induced in anesthetized 40 kg domestic pigs. Defibrillation was attempted after either 4 or 7 min of untreated VF. Failing to reverse VF, a I min interval of precordial compression and mechanical ventilation preceded each subsequent defibrillation attempt. The amplitude frequency spectrum of digitally filtered VF wavelets was computed with Fourier analysis during uninterrupted precordial compression from conventional right infraclavicular and left apical electrodes. Of a total of 34 electrical defibrillation attempts, 24 animals were restored to spontaneous circulation (ROSC). An amplitude spectrum analysis (AMSA) value of 21 mV Hz had a negative predictive value of 0.96 and a positive predictive value of 0.78. Conclusions: AMSA predicted when an electrical shock failed to restore spontaneous circulation during CPR with a high negative predictive value. This method potentially fulfills the need for minimizing ineffective defibrillation attempts and their attendant adverse effects on the myocardium. (C) 2002 Elsevier Science Ireland Ltd. All rights reserved.