The Real-time Quality Analysis Method Of Pulse Wave Signal And Its Engineering Realization

In the research and application of pulse wave detecting equipment, the reductionof signal quality caused by interference and motion artifact has always been aproblem of researchers. Vulnerable to inside or outside interference, pulse wave signalis unable to provide accurate information about cardiovascular function of clinicalvalue. Farther more, the detecting person must have certain knowledge andexperience about pulse wave detection. All of these become difficult factors for pulsewave in its family care and remote medical application.In order to evaluate the pulse wave signal quality during the acquisition, awaveform standardization and normalization method base on pulse wave formcharacteristics is involved. The method can improve the accuracy of cardiovascularfunction information and the degree of automation. The pulse wave formcharacteristics are quantitative described as pulse wave form characteristicsparameters. Considering the physiological significance of those characteristicparameters, pulse wave form can be classified with the reference of large number ofclinical pulse wave.Using it as a base, a real-time pulse wave detection quality evaluation model isestablished. In order to solve the practical engineering requirements of real-timeanalysis of pulse wave signal acquisition quality, real-time single-wave discriminatetechniques, including pulse wave single-wave discriminate algorithm in processingqueue, experience heart rate preliminary estimation algorithm and real-timesingle-wave discriminate algorithm based on experience in heart rate is designed.Parameters and those algorithm evaluated pulse wave quality are introduced,including pulse frequency energy ratio, interference frequency variance ratio, baselinedrift accumulation, cycle variance, single-wave data point feature variance, singlewave main peak width variance. Real-time pulse wave collecting quality evaluationmodel is established through a variety of means and methods，and a logical validationis made.In terms of equipment development, a real-time analysis system based on ARMCORTEX A8device is designed. The system uses Samsung S5PV210processor CPUcore and peripheral expansion circuit board combinations. Samsung S5PV210core isthe main processor, using the2.0mm distance between double row needles which leads to the most commonly used functions CPU pins. Core board hardware resourcesalso include512M DDR2memory and1GB of flash memory (SLG). The design ofperipheral circuit part mainly includes the AD sampling circuit, sound card circuit,LCD circuit, USB, network interface circuit, power circuit, etc. A real-time pulsewave quality evaluate analysis software base on Android platform is developed. Theoverall structure is divided into six layers, namely sensing layer, presentation layer,business logic layer, data access layer, service layer, network-based communicationlayer. Its main functions cover AD acquisition, sound card acquisition, real-timequality assessment, SQLite database functions and network communication functions.Finally, a double blind experiment is designed to evaluate the effectiveness of themodel and equipment. With the analysis of reproducibility and variability ofexperimental data, discussion about effectiveness of the model under condition ofinduced disturbance is made.