基于随机森林模型的动态动脉血管硬化指数估计方法  

Research on ambulatory arterial stiffness index estimation using random forest model

作　　者：张海康 程云章[1] 张天逸[1] ZHANG Haikang;CHENG Yunzhang;ZHANG Tianyi(Shanghai Interventional Medical Device Engineering Technology Research Center,University of Shanghai for Science and Technology,Shanghai 200093,China)

机构地区：[1]上海理工大学上海介入医疗器械工程技术研究中心,上海200093

出　　处：《生物医学工程研究》2022年第1期55-61,共7页Journal Of Biomedical Engineering Research

基　　金：上海工程技术研究中心资助项目(18DZ2250900)。

摘　　要：针对目前检测动脉硬化程度需要昂贵的设备以及专业操作人员的现状,本研究提出了基于机器学习的动态动脉硬化指数(ambulatory arterial stiffness index,AASI)估计方法。从UCI Machine Learning公共数据库获取患者的光电容积脉搏波信号、心电信号以及动脉血压信号,将其预处理后求得作为机器学习模型的特征值。使用随机森林模型进行AASI的估计,并将最优的回归模型送入遗传算法(genetic algorithm,GA)筛选最优特征子集,完成AASI的估计。实验结果表明,经过GA的特征子集优化,随机森林模型的MAE从0.0974减少至0.0855。对所得结果进行统计学分析后,发现本研究方法针对高血压患者时表现不佳,后续将进一步优化模型,以提高性能。Considering the expensive devices and experienced doctors during the measurement of arterial stiffness,we presented a novel method based on ambulatory arterial stiffness index(AASI).The raw datas included of photoplethysmography,electrocardiography and arterial blood pressure were obtained from UCI Machine Learning dataset.The features of machine learning were used as inputs of machine learning model after preprocessing and feature extraction.To explore the optimal feature set,genetic algorithm(GA)was employed.With the help of GA,RF′s MAE was reduced from 0.0974 to 0.0855.After the statistical analysis of the results,we found that this method performed poorly to patients with high blood pressure value.We will further optimize the model to improve the performance in the future.

关 键 词：生物医学信号 特征提取 择优算法 RR间期 局部加权回归 

分 类 号：R318[医药卫生—生物医学工程] O244[医药卫生—基础医学]