Fundamental study of measurement of cardiorespiratory signals in a sitting position using piezoelectric sensors

In this paper, we propose a measurement system to extract the heartbeat and respiration signals from piezoelectric sensors located in a chair seat. Eight healthy male subjects (22-26 years old) sat on a chair with piezoelectric sensors. The experiment consisted of five acquisitions (300 s recording with 30 s resting). In the 1st acquisition, the subject was instructed to maintain a relax state. During the next three acquisitions, the subject maintained respiration rates of 3, 4, and 5 s. The order was randomly given for each subject. Finally, the subject was instructed to stay in a resting state for 300 s. Band-pass filters were used to separate the heartbeat and respiration signals from the output signal of the piezoelectric sensors (heartbeat: 0.7-7 Hz, respiration: 0.1-0.4 Hz). Then, the standard heartbeat interval was calculated using the autocorrelation function to generate a template. Heartbeat and respiration signals were extracted using a two-dimensional cross-correlation pattern matching method. For validation, the error ratio between the piezoelectric sensor data and the reference data was determined. The error ratios of the heartbeat interval and respiration interval were 2.89±4.41% and 5.57±4.42%, respectively. Therefore, it is suggested that heartbeat and respiration signals extracted from piezoelectric sensors in a sitting position can be used as an alternative method for extracting biological signals.