The Research On Automatic Diagnosis Technology Of ECG Signal Based On FPGA

With the improvement of living standards and accelerated pace of life,the Ineidence of cardiovascular disease has become a main threat to human health.With the characteristics of sudden attack,It is impossible to be supine in the hospital for a long time,so the real-time health monitor work is very needed,therefore,researching high-performance and portable ECG monitor has great significance. With the background of the existing medical,social,science and technology,the field of medical care has been developed toward personalization,portability,multifunction.For the above,this paper put forward a ECG monitoring system solutions based on SOPC,Altera corporation CycloneII series chip EP2C20 as the core chip of FPGA hardware platform. System hardware design include ECG singal per-processing module and Functional modules within the FPGA. Nios II embedded soft core as the system processor,System software design is mainly carried out in the NiosII IDE, using the NiosII custom instruction methods to compile the C language applications. Applications include the SD card storage procedures based on FAT file system,LCD display control program,A/D,D/A converter control procedures.ECG signal pre-processing module includ analog filter and digital filter.Analog filter is mainly completed by the ECG conditioning module, including the ECG signal extraction circuit,filter circuit,amplification circuit. the amplified ECG signal bias to the input range of A/D converter,restraining frequency interference to some extent.For the digital filters,using Altera provided IP core to design,firstly taking functional simulation in DSP builder,and verifying the FIR filter through the data in the MIT-BIH database.We added all the functional modules which have been achieved and other external interfaces to the Avalon Bus,so that they can constitute an embedded system based on NiosII processor.Through it,we can make use of the flexible function of NiosII processor to complete the whole system workflow control of data acquisition,transmission,storage,communication,display,etc.Various existed algorithm for ECG autoamted diagnosis at home and abroad in recent years are summarized in this paper.using support vector machine for processing ECG data with building classification model,training classifier,diagnosising coronary heart disease. The algorithmis modeled and simulated in Matlab software,and verified with UCI coronary heart disease database.using the Grid optimization algorithm to find classifier optimization parameters, finally classifier obtain the highest classification accuracy.