基于脉搏波传导时间和脉搏波特征参数的连续血压无创检测

Non-invasive measurement of continuous blood pressure based on pulse transit time and pulse wave characteristic parameters

本研究为克服在基于脉搏波传导时间(pulse transit time,PTT)方法无创血压检测中个体差异对测量准确性的影响,分析了个体脉搏波特征参数与血压值的相关性,并将个体脉搏波特征参数中与血压值相关度高的参数作为优化脉搏波特征参数加入PTT与血压的校正模型中,以提高模型普适性。基于偏最小二乘法(partial least squares,PLS)对50名志愿者200组PTT和脉搏波特征参数数据进行训练建模,得到舒张压、收缩压的预测模型。再选取5名新志愿者的PPT和脉搏波特征参数进行预测,最大预测误差小于5 mm Hg,满足AAMI国际电子血压计标准。将优化的脉搏波特征参数和PTT引入到连续血压预测模型当中,有助于提高血压预测模型的准确度和普适性,有助于无创连续血压检测的临床应用。

For individual difference,the correlation between the characteristic parameters of pulse wave and pressure value were analyzed,and the correlation model of blood pressure and PTT to improve the accuracy of the prediction of continuous blood pressure was established.Based on partial least squares(PLS),the training constructing mode was done by collecting PTT and characteristic parameters of pulse wave of 200 groups of 50 volunteers,and the prediction model of diastolic blood pressure and systolic blood pressure was established.We selected randomly the PPT and the characteristic parameters of pulse wave of 5 new volunteers and made predictions.The maximum prediction error was less than 5 mmHg,which met the standard of AAMI international electronic sphygmomanometer.The optimized characteristic parameters of pulse wave and PTT are introduced into the prediction model of continuous blood pressure,which is helpful for improving the accuracy and universality of BP prediction model and the clinical application of noninvasive continuous blood pressure measurement.