Heart rate reliability prediction for the estimated value from sheet-shaped microdisplacement sensor signal

Monitoring heart rate has possibility to detect signs of cardiovascular diseases such as hypertension because these diseases are related to abnormal autonomic nervous function that has a relationship with heart rate variability (HRV). Some previous studies have been proposed estimate methods of HRV based on signals from non-contact measurements, however, there is few techniques to predict the reliability of the estimated value without the true heart rate signal. This paper proposes a new method to estimate HRV of a person in a chair using signals from sheet-shaped microdisplacement sensors attached on the chair. In addition, a technique to predict the reliability of the estimated HRV is introduced to know whether the estimated value is reliable or not even under conditions difficult to obtain information of real HRV. An experiment 27 subjects participated in was carried out to evaluate the performance of the proposed method. Experimental results showed that the estimate accuracy, which is defined as a correlation coefficient between true and estimated HRVs, is 0.87. The accuracy of the proposed method is higher than that of a previous study.