Abstract
Data in recordings obtained from ambulatory patients using wearable sensors are often corrupted by motion artefact and are, in general, noisier than the data obtained from the nonmobile patients. Identifying and ignoring erroneous measurements from these data is very important, if wearable sensors are to be incorporated into clinical practice. In this paper, we propose a novel Signal Quality Index, intended to assess whether reliable heart rates can be obtained from a single channel of ECG collected from ambulatory patients, using wearable sensors. The proposed system is based on wavelet entropy measurements of the heart rate variability signal. The system was trained and tested on expert-labeled data from a particular wearable sensor and was also tested on labeled data from a different sensor. The sensitivities and specificities achieved were 94% and 98%, respectively, on data from the same sensor as the training set, and 91% and 97%, respectively, on data from a different sensor, indicating the potential of the system to generalize across different sensors. Because the system relies on a single channel of ECG, it has the potential for inclusion in applications using wearable sensors and in the most basic clinical environments.
Highlights
T HERE is a widespread consensus that wearable sensors will be a key part of delivering healthcare in the future
We propose a new algorithm for classifying segments of ECG as “acceptable” or “unacceptable” (for obtaining reliable heart rate (HR) measurements), which is based on spectral analysis of the heart rate variability (HRV) signal
We investigated the performance of quadratic, polynomial, and radial basis function (RBF) kernels, and obtained the best performance using a RBF kernel, which is given by k (x, y) = exp − x − y 2 /2σ2 (8)
Summary
T HERE is a widespread consensus that wearable sensors will be a key part of delivering healthcare in the future. The issue of identifying unreliable data is important, as data obtained from ambulatory patients, the patients more likely to benefit from the use of wearable sensors, are more likely to contain artefact than data obtained from bed-bound patients [1]. The electrocardiogram (ECG), routinely collected from hospital patients, is often contaminated with noise leading to unreliable vital sign measurements. Erroneous vital sign measurements may result in a large number of false alerts that can lead to the phenomenon of “alarm fatigue,” whereby ward staff become desensitized to and ignore alerts from the monitoring. Manuscript received April 8, 2016; revised August 12, 2016 and August 31, 2016; accepted September 30, 2016. Date of publication October 5, 2016; date of current version September 1, 2017
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
