PulSync: The Heart Rate Variability as a Unique Fingerprint for the Alignment of Sensor Data Across Multiple Wearable Devices

Abstract

Most off-the-shelf wearable devices do not provide reliable synchronization interfaces, causing multi-device sensing and machine learning approaches, e.g. for activity recognition, still to suffer from inaccurate clock sources and unmatched time. Instead of using active online synchronization techniques, such as those based on bidirectional wireless communication, we propose in this work to use the human heartbeat as a reference signal that is continuously and ubiquitously available throughout the entire body surface. We introduce PulSync, a novel approach that enables the alignment of sensor data across multiple devices utilizing the unique fingerprint-like character of the heart rate variability interval function. In an evaluation on a dataset from 25 subjects, we demonstrate the reliable alignment of independent ECG recordings with a mean accuracy of −0.71 ± 3.44 samples, respectively −2.86 ± 11.43 ms at 250 Hz sampling rate.

Agreement between two photoplethysmography-based wearable devices for monitoring heart rate during different physical activity situations: a new analysis methodology

“…This methodology aims to avoid the interpolation of the HR data over longer average times, and instead advocates for a comparison over different time scales to track variations of HR along recording. Interestingly, the proposed synchronization procedure does not require any interpolation of time series, in contrast to recently proposed methods such as in 32 , which promote resampling of time series for delay estimation. In Appendix A4, a study is presented assessing the effects of interpolation using the proposed methodology versus the different time series at 25 Hz as in 32 ).…”

“…Interestingly, the proposed synchronization procedure does not require any interpolation of time series, in contrast to recently proposed methods such as in 32 , which promote resampling of time series for delay estimation. In Appendix A4, a study is presented assessing the effects of interpolation using the proposed methodology versus the different time series at 25 Hz as in 32 ). The study shows differences in LoA lower than 2 bpm, suggesting it can be of importance only when the compared systems show good agreement.…”

Timing errors and temporal uncertainty in clinical databases—A narrative review

“…Artifacts or other information common to both signals can be identified and retrospectively aligned in situations where system constraints preclude real-time synchronization. This process may rely upon artifacts that are introduced into the data streams while they are being recorded ( 201 , 209 – 211 ), or may utilize features such as heart beats that are naturally present and accessible in a wide range of high frequency physiological signals ( 212 – 216 ). This method of synchronization requires artifacts or features from multiple time points throughout the signals to ensure that clock drift is properly accounted for ( 85 ).…”

IBSync: Intra-body Synchronization of Wearable Devices Using Artificial ECG Landmarks

“…The "naturally synchronized" heartbeat has also been used in a time division multiple access (TDMA) protocol for medium-access control (MAC) [29]. In our previous research on PulSync, the heart rate variability (HRV) serves as a unique fingerprint to align sensor recordings from different wearable devices [53].…”

“…Originated in activity recognition, there exist methods to align measurements offline, after the recording [1,5,8,9,50]. The used gestures and motion patterns are, however, not incidental but rather tend to be cumbersome, obtrusive, and suffer from inaccuracies due to soft tissue deformation and delays due to motion sequences and inertia of the body parts [30,53].…”