Study Of Pulse Taking Depth Localization And Pulse Signal Processing Based On Multi-parameter Fusion

The change in pulse waveform characteristics is an important basis for evaluating the physiological and pathological status of the human cardiovascular system.But the existing measurements are operated by rigid mechanisms and mostly use pressure as the only indicator of pulse picking to find the best position,relying on experience to determine the best location of the pulse picking point,which is time-consuming and labor-intensive and has low positioning accuracy,and cannot accurately adapt to individual differences.Therefore,it is important to study the pulse pickup depth positioning to meet the depth calibration of different individuals and to achieve accurate measurement of the full waveform.In this paper,we propose a method to locate the optimal pulse-taking point using the first-order derivative of contact stiffness and combine it with cardiovascular parameters to improve the localization accuracy,and propose a signal processing algorithm with a progressive semi-soft thresholding method combined with Dohono threshold estimation for pulse noise interference.The main elements are as follows.(1)By constructing a three-dimensional quasi-static radial artery finite element model,the effects of different soft tissue thickness and vascular elasticity on contact stiffness are analyzed.the mechanism of vascular compression and deformation is elucidated,and the location of the optimal pulse-taking point is determined.The amplitude change law under the action of variable stiffness indenter is studied,and the simulation results are useful as a guide for locating the optimal pulse-taking depth.(2)A pulse acquisition analysis system for depth calibration is built.Combined with the contact stiffness calculation needs,the pressurized structure of the stepping slide and the piezoresistive pulse sensor is designed.Self-developed chips were used to realize signal amplification,acquisition and A/D conversion,and STM32 minimal system board was used to realize motor control as well as upper computer communication functions.The Lab VIEWbased upper computer interface is designed to realize the control,signal acquisition and analysis,and data storage of the stepper slide.The system features high motion accuracy,good antiinterference degree,and high integration.(3)Pulse signal processing research.An improved asymptotic semi-soft threshold function combined with Dohono threshold estimation algorithm is proposed,and different wavelet functions,decomposition scales and commonly used threshold estimation methods are compared and analyzed using signal-to-noise ratio,standard deviation and smoothness,and the results show that the algorithm proposed in this paper has the optimal effect for removing disturbances such as noise and vibration.Feature extraction is performed on the processed signal to facilitate subsequent experimental analysis.(4)Experimental study of pulse-taking depth localization method.Through the analysis of contact stiffness influencing factors,the pulse picking loading strategy based on stiffness prediction is proposed.The validity of the first-order guide of stiffness to locate the optimal pulse pickup point is verified,and a method to enhance the depth localization accuracy using the cardiac output and pulse eigenvalues is proposed.The applicability of depth localization is further verified by simulating Chinese medicine techniques and clarifying the change pattern of pulse wave characteristics when changing the pulse pickup angle and indenter stiffness.In summary,this paper proposes an improved wavelet threshold denoising algorithm and a depth localization method with multi-parameter fusion,and designs a corresponding pulse and stiffness information acquisition system,and verifies its effectiveness through experiments,and the theoretical results can provide new ideas for the objectification of pulse diagnosis in TCM.