Nonlinear Analysis On Human Pulse And Research On Effects Of Different Emotions And Environmental Conditions

Pulse waves play really important role in the research field of cardiovascular system,which directly reflect the fluctuation of heart and artery cardiovascular system. In this paper, thegenerating physiological mechanisms of pulse waves are firstly particularly described according tothe view of physiological anatomy and nervous system.An efficient de-noising algorithm for removing baseline drift, disturbances of workingfrequency and electromyography in human pulse signals is proposed based on wavelet transformtheory. Then the de-noised pulse signals are decomposed using wavelet decomposition and certaindetail components are reconstructed. The peaks of main pulse wave, i. e. P-waves, are extractedbased on the above reconstructed detailed signals and threshold approach firstly. And then, fivefeature points of the pulse signal are identified. The extraction of T-wave peaks, D-wave peaks,V-wave troughs and starting points are achieved depending on the position relations between thesepoints and P-wave peaks.The continuous pulse waves exhibit obvious periodically and with discrete frequency spectra.The amplitude frequency spectra are calculated for different physiological states here. Pulse wavesignals also show complicate multi-frequency band characteristics. The time-frequency changingrules of pulse waves in different physiological states are analyzed with a given wavelet packagebased energy ratio method.The time series of the pulse main peaks (i. e. PP wave intervals) are similar with HRVs.Nonlinear characteristics of PP intervals are mainly analyzed for different physiological states. Withsurrogate data method, which is applied to identifying the existing chaos of PP intervals of pulse, itis proved that the series of PP intervals of pulse are chaotic. After introducing details of the surrogatedata method, some characteristic parameters in the method are used to compare differentphysiological and pathological states.Some nonlinear parameters of PP wave intervals are calculated here, including correlationdimensions, maximum Lyapunov exponents, approximate entropy and complexity. All these valuesare used to distinguish quantitatively for the differences among physiological and pathologicalstates. Then a digital signal measuring and recording device for pulse waves is developed whichincludes local flash memory and multi-communication channels, in which suitable amplifier, filterand advanced embedded system techniques are adopted carefully. The really weak pulse waves atfinger can be detected easily, conveniently and accurately with it.At last, the changes of pulse waves are studies for different emotion stress states by usingnonlinear analysis methods. Furthermore, some research on the changes of pulse waves undersituation of vibration and noise environment in a car are also studied.In summary, nonlinear characteristics of pulse waves are analyzed based on measured data.With nonlinear dynamics, chaos identification theory nonlinear estimation together with wavelettransform technology, pulse wave signals are treated for de-noising and extracting of characteristicpoints. Nonlinear description and identification are finished for different pulse waves in differentphysiological status, such as healthy, sexual, age, sleeping states etc. Furthermore, these also areapplied on the research under different emotions and the special situation of car vibration and noiseenvironment.Certainly, there are much more things to do to explain the principles of pulse waves in differentphysiological statuses, and the algorithms and some conclusions also need much more clinicalexperiment to be testified.