Respiratory activity is the most important basic life activity of human body, respiratory state detection is of great practical significance. Based on the characteristic of high correlation between the change of tidal volume and the change of abdominal displacement, a method of detecting respiratory status through abdominal displacement data is proposed. The method uses a gas pressure sensor to collect the tidal volume in the steady state of the subject once, which is used as the baseline data. The abdominal displacement data of the subject in the three breathing states of slow breathing, steady breathing and rapid breathing were collected with an acceleration sensor. The warp path distance between the lung and abdominal data in the three different states was calculated, this warp path distance together with the period extracted from the abdominal data is used as a two-dimensional feature and input to the support vector machine classifier. The experiments show that the accuracy of the classification results reaches 90.23%. The method only needs to measure the lung data once in smooth breathing state, and the subsequent continuous detection is achieved by measuring the displacement of the abdomen only. This method has the advantages of stable and reliable acquisition results, low implementation cost and simplified wearing method, and has high practicality.
Keywords: Correlation; Respiratory cycle; Respiratory state detection; Support vector machine; Warp path distance.
©2023PublishedbyElsevierLtd.