Research And Application Of Respiratory Parameter Extraction Based On Angular Velocity Signal

Respiratory physiological parameter is an important reference for people to understand their own situation.With the improvement of health awareness,people expect long-term respiratory monitoring,which puts forward higher requirements for respiratory signal acquisition devices.As a kind of new technology based on inertial sensor which can be used in intelligent wear,human body sensor network and respiratory detection,it has been widely studied and explored in recent years.However,among all the existing studies,there has been no study about single channel respiratory signal classification and identification.The inertial sensor includes the accelerometer and the angular velocity meter.Based on the advantages of the angular velocity meter in obtaining high quality respiratory signal,angular velocity meter is chosen as the tool for obtaining respiratory angular velocity.This paper validates the feasibility of extracting respiratory signals from single channel respiratory angular velocity,and then studies the application of the respiratory angular velocity in extracting respiratory phase and respiratory signal classification.The main work of the study can be divided into two parts: analysis and application.For the analysis,this paper is divided into two parts: the signal acquisition and analysis of the breathing parameters.A signal inertial sensor is placed in the sternal notch to obtain the respiratory angular velocity signal.The single channel device can facilitate the collection of respiratory signal and sternal notch position can ensure the robustness of the respiratory signal.The respiratory rate and the respiratory phase are used as verified parameters and the end-tidal carbon dioxide is used as the reference respiratory signal for comparison.The results show that the respiratory rate is within the confidence interval and the phase shift median error is less than 0.5 second.For the application,firstly,respiratory phases are extracted from the respiratory angular velocity by converting it into an easily recognizable respiratory angle signal.In addition,a rehabilitation therapy instrument is designed,which is one of the application of the respiratory phase and one of the workload of this paper.Secondly,the respiratory signals are classified and seven kinds of common respiratory abnormalities and normal respiratory signals are selected.Based on the previous research,the Support Vector Machines(SVM)is used to design the classifier for pattern recognition.In the design of the classifier,a variety of techniques are used to obtain a variety of better eigenvalues.The eigenvalues include: mean,variance,energy,overthreshold respiratory number and Symbolic Aggregate Approximation(SAX).In order to improve the validity of eigenvalues,Wavelet technology and window segmentation technology are used.The best accuracy of classification is 91.25% by ten-fold cross validation,which validate the feasibility of using single channel respiration angular velocity signal to classify respiratory signals.In summary,the results of this study show that the single-channel breath angular velocity signal collected by inertial sensor can replace the traditional respiratory detection instrument to obtain the respiratory rate and obtain a more accurate respiratory phase.The signal also can be used to extract the respiratory phase and classification of respiratory waveforms has a good classification effect.The work of this paper presents a new way of breathing detection,which provides a reliable idea for long-term respiratory detection and can be applied to early warning of respiratory-related diseases.