基于SOPC的心电信号QRS波和ST段检测与识别  

Detection and Analysis of QRS Wave and ST Segment of Electrocardiosignal Based on SOPC

作　　者：李在军[1] 王明[1] 钟维[1] 黄启俊[1] 常胜[1] 王豪[2] 

机构地区：[1]武汉大学物理科学与技术学院,湖北武汉430072 [2]武汉大学微电子与信息技术研究院,湖北武汉430072

出　　处：《武汉大学学报（理学版）》2014年第2期122-128,共7页Journal of Wuhan University:Natural Science Edition

基　　金：湖北省自然科学基金(2011CDB272)资助项目

摘　　要：采用小波变换法和一种改进的基于状态机的极值对过零点检测算法在FPGA(现场可编程逻辑门阵列)硬件平台上实现了心电信号QRS波的检测和ST段特征点的提取,然后采用BP(反向传播)神经网络的方法在FPGA嵌入的软核niosII的软件平台上实现了对ST段的三种形态识别.通过小波变换和极值对过零点检测算法提取的ST段的数据作为BP神经网络的测试输入.用MIT-BIH数据库中的心电信号数据作为数据来源,选取Altera公司的FPGA开发板DE-2(EP2C35F672C6)作为验证平台,结果 R波峰检测和ST段的形态识别的正确率分别达到97.4%和96.8%,硬件资源共消耗4 309个逻辑单元,系统的整体延时时间为0.52s.结果表明该设计方案可以满足心电信号的实时分析.An SOPC（system on programmable chip）based detection and analysis of electrocardiosignal is designed and implemented to meet the demand for the need of intelligentized,miniaturized and real-time electrocardiogram equipment.The detection of QRS wave and ST segment feature point is realized with the method of wavelet decomposition and a improved arithmetic based on state machine which is presented to test the zero of the extremum couple on the platform of FPGA（field programmable gata array）,and then the feature point of ST segment are picked up,and the pattern recognition of ST segment is realized with the method of BP（back propagation）neural network on the plat form of embedded soft-core NIOSII.The altera FPGA Development Board DE-2（EP2C35F672C6）is chosen as the verification platform and the ECG data of the MIT-BIH arrhythmia database is chosen as the source of data.The correct rate of R wave crest and the recognition of ST segment come to 97.4%and 96.8%separately.The hardware resource consumption is 4 309logic elements and the delay of the whole system is 0.52s.The result indicates that the designing scheme that combine software and hardware based on sopc can meet the request of real time analysis of electrocardiosignal.

关 键 词：小波变换 QRS波 ST SOPC BP神经网络 

分 类 号：TN919[电子电信—通信与信息系统]