Adaptation Analysis And Feature Extraction Of Neural Electrical Signal Sevoked By Manual Acupuncture

Acupuncture is an important part of Chinese medicine with unique ways of thinking and richexperiences whose efficacy has been proved by Chinese clinical research during thousands ofyears. However，it is unclear whatthe exact mechanism of acupuncture is and how acupuncturehas effect on the neural system. The neural system plays a leading role within the organism.Acupuncture can evoke the nervous system to various kinds of neural electrical signals. In thethesis, modern signal processing techniquesanalysis methods are applied to analyze thecharacteristics of neural electrical signals, which evoked by different manual acupuncture (MA)from peripheral neural system and electroencephalogram(EEG) in order to explore the effect ofacupuncture on the neural system.Spike-frequency adaptation is a common property of neurons and plays an important role inneural information processing. EEG and brain function are inseparable, because the brain isbody’s processor. We design experiments about Sprague-Dawley (SD) and human body toexplore the acupuncture neural system conduction and effect. Acupuncture at Zusanli acupointwith four different MA frequencies to obtain action potentials at dorsal spinal nerve root fromperipheral nervous system in SD ratandobtainEEGfrom high central neural system in humanbody. The main content includes two parts as follows:Firstly, for action potential sequences at dorsal spinal nerve root, we introduce the concept ofthe interspike interval (ISI) to analysis spike-frequency adaptation. A modified Hodgkin-Huxleytype neuron model with an additional adaptation AHP-current is taken into account to study thefiring pattern, and compared with experiment results evoked by manual acupuncture in our study.Secondly,the concept ofanalysis methods areintroduced, such as power spectrumanalysis,Detrended Fluctuation Analysis,Approximate Entropyand Sample Entropytoextractcharacteristic parametersof the acupuncture EEG.The power spectrum estimation basedon Burg algorithm is used to analyze the changes of brain electric energy. Detrended FluctuationAnalysisis analyzed the volatility of MA EEG. To measure complexity of MA brain electricalsignals, two time series nonlinear analysis such as Approximate Entropyand Sample Entropy areapplied.The obtained results can give some theoretical supports to reveal the acupuncture actionmechanismbased on neural electrical information.