Automatic detection of left ventricular ejection time from a finger photoplethysmographic pulse oximetry waveform: comparison with Doppler aortic measurement
Automatic detection of left ventricular ejection time from a finger photoplethysmographic pulse oximetry waveform: comparison with Doppler aortic measurement
复制标题
DOI：
10.1088/0967-3334/28/4/009
复制
发表时间：
2007-04-01
影响因子：
3.2
通讯作者：
Lovell, Nigel H.
中科院分区：
文献类型：
--
作者：
Chan, Gregory S. H.;Middleton, Paul M.;Lovell, Nigel H.
左心室射血时间（LVET）是心室性能和前负荷的有用量度。本研究探讨了一种新的方法，连续LVET监测使用无创手指光电容积脉搏血氧波形（PPG-POW）。提出了一种基于导数分析、波形平均和基于规则的逻辑相结合的从手指PPG-POW中自动逐拍检测LVET的方法。在13名健康受试者分级头高位倾斜的检测方法的性能进行了评价。总体而言，PPG-POW衍生的LVET与主动脉血流衍生的LVET之间的相关性较高且显著（r = 0.897，p < 0.05）。偏倚为-14 +/- 14 ms（平均值+/- SD），误差百分比为9.7%。尽管这些结果不足以满足要求绝对准确性时LVET临床评价的要求，但受试者内PPG-POW LVET和主动脉LVET之间的强相关性（r = 0.945 +/- 0.043，平均值+/- SD）将支持应用PPG-POW检测个体LVET的方向变化。这可能是非常有用的早期识别进行性低血容量症或失血。本研究已经证明了一种很有前途的方法，从一个非侵入性的，易于获得的信号，可以很容易地从现有的病人监测获取潜在有用的信息是必要的，以评估本方法在临床护理监测的适用性。
Left ventricular ejection time (LVET) is a useful measure of ventricular performance and preload. The present study explores a novel method of continuous LVET monitoring using a noninvasive finger photoplethysmographic pulse oximetry waveform (PPG-POW). A method for the automatic beat-to-beat detection of LVET from the finger PPG-POW is presented based on a combination of derivative analysis, waveform averaging and rule-based logic. The performance of the detection method was evaluated on 13 healthy subjects during graded head-up tilt. Overall, the correlation between the PPG-POW derived LVET and the aortic flow derived LVET was high and significant (r = 0.897, p < 0.05). The bias was - 14 +/- 14 ms (mean +/- SD), and the percentage error was 9.7%. Although these results would not be sufficient to satisfy the requirement for clinical evaluation of LVET when absolute accuracy was demanded, the strong correlation between the PPG-POW LVET and the aortic LVET on an intra-subject basis (r = 0.945 +/- 0.043, mean +/- SD) would support the application of PPG-POW to detect the directional change in LVET of an individual. This could be very useful for the early identification of progressive hypovolaemia or blood loss. The present study has demonstrated a promising approach to extract potentially useful information from a noninvasive, easy-to-obtain signal that could be readily acquired either from existing patient monitoring is necessary to evaluate the applicability of the present approach in clinical care monitoring.