Interval-Level Validation of a Wearable Dry-Electrode ECG Biosensor Platform: A Proof-of-Concept Study

Abstract

Wearable dry-electrode electrocardiographic (ECG) systems may enable longer and more comfortable rhythm monitoring, but their interval-level measurement fidelity requires validation against reference acquisition systems. In this single-center proof-of-concept study, 20 participants (10 healthy volunteers and 10 cardiology patients) underwent simultaneous ECG recordings with a U.S. Food and Drug Administration (FDA)-cleared wearable dry-electrode platform (X-trodes System M) and a wired reference system (MP150 with ECG100C module, Biopac Systems). Three-minute paired segments were recorded and underwent temporal synchronization, bandpass filtering, baseline correction, automated discrete-wavelet peak detection in NeuroKit2 (version 0.2.10), and standardized manual adjudication. After predefined exclusions for temporal alignment and P-wave detection quality, agreement was assessed in nine cardiac and ten healthy participants for RR intervals and in seven cardiac and eight healthy participants for P-to-R peak intervals. RR interval agreement was excellent in the healthy and cardiac cohorts, with a mean bias of 0.4 ± 3.1 ms and 0.1 ± 4.6 ms and mean absolute error (MAE) of 0.97 ms and 1.28 ms, respectively. Intraclass correlation coefficient (ICC) was over 0.99 for both cohorts. P-to-R peak interval agreement was strong, with an MAE of 1.76 ms and 7.14 ms and ICC of 0.96 and 0.98, respectively. These findings provide preliminary evidence of interval-level agreement for the X-trodes wearable dry-electrode ECG biosensor platform under controlled, short-duration daytime conditions. Because the matched interval pairs were clustered within a small number of participants, the findings should be regarded as proof-of-concept evidence supporting larger ambulatory validation studies rather than as definitive clinical validation.