基于SVM回归的连续血压测量方法

针对目前血压测量主要以间歇式为主, 且难以摆脱充气袖带束缚的问题, 在脉搏波特征参数法测量血压的基础上, 提出基于SVM(Support Vector Machine)回归的连续血压测量方法。以脉搏波为主要研究对象, 提取脉搏波相关特征点, 并对其进行周期分割, 得到脉搏波时域特征, 寻找动脉血压和脉搏波时域特征之间的关系, 建立连续血压回归模型。实验结果表明, 该方法可以很好地对每一心拍的血压值进行连续测量, 且测量结果符合美国医疗器械促进学会(AAMI: Association for the Advancement of Medical Instrumentation)推荐的标准,平均误差不超过0. 67 kPa、标准差不超过1. 07 kPa。

Continuous Blood Pressure Measurement Method Based on SVM Regression

The current clinical measurement of BP(Blood Pressure) is mostly based on the intermittent type, and cannot get rid of the shackles of the inflatable cuff. On the basis of measuring blood pressure by pulse wave characteristic parameter method, a measurement based on SVM(Support Vector Machine) regression is proposed to realize continuous blood pressure measurement. Taken the pulse wave for the main object of study and extract their feature points, divided the pulse wave according to these feature points, then the time domain characteristics of pulse wave were obtained. To find the relationship between arterial blood pressure and these characteristics, a regression model based on SVM was established. The experimental results show that this method can measure continuous blood pressure well, the average error of BP measurement are all less than 0. 67 kPa, and the standard error are all less than 1. 07 kPa, this can meet the standards of the AAMI(Association for the Advancement of Medical Instrumentation).