文章摘要

文章信息/Info

Title:: Study on blood pressure detection method based on light gradient boosting machine model

Author(s):: WU Shaowu; XU Yuru; College of Mathematics and Science, Beijing University of Technology, Beijing 100124, China

Keywords:: Blood pressure; Pulse wave; Feature extraction; Light gradient boosting machine model; Cross-validation

摘要:: 为了提高血压的检测精度，本研究基于光电容积脉搏波（photoplethysmography，PPG）波形，提取该波形的时域特征参数15个，并在此基础上，加入前一时刻的血压值作为特征，通过LightGBM（light gradient boosting machine,LightGBM）模型进行10折交叉验证。结果表明，舒张压和收缩压的平均绝对误差均小于5 mmHg，均方根误差均小于8 mmHg，满足了AAMI国际电子血压计标准。本研究所用的检测方法简单高效、易于实现，有助于个性化血压检测的应用以及临床应用。

Abstract:: In order to improve the accuracy of blood pressure detection,we extracted 15 time-domain characteristic parameters of the waveform based on photoelectric volume pulse wave (PPG) waveform. On this basis, the blood pressure value at the previous moment was added as the feature, and the LightGBM model was used for 10-fold cross-validation. The results showed that the mean absolute error of diastolic blood pressure and systolic blood pressure was less than 5 mmHg, and the root mean square error was less than 8 mmHg, which met the standard of AAMI international electronic sphygmomanometer. The method used in this paper is simple, efficient and easy to implement, which is helpful to the application and clinical application of personalized blood pressure detection.