Methods for the metrological characterization of wearable devices for the measurement of physiological signals: state of the art and future challenges

Abstract

Wearable devices are rapidly spreading in many different application fields and with diverse measurement accuracy targets. However, data on their metrological characterization are very often missing or obtained with non-standardized methods, hence resulting in barely comparable results. The aim of this review paper is to discuss the existing methods for the metrological characterization of wearable sensors exploited for the measurement of physiological signals, highlighting the room for research still available in this field. Furthermore, as a case study, the authors report a customized method they have tuned for the validation of wireless electrocardiographic monitors. The literature provides a plethora of test/validation procedures, but there is no shared consensus on test parameters (e.g. test population size, test protocol, output parameters of validation procedure, etc.); on the other hand, manufacturers rarely provide measurement accuracy values and, even when they do, the test protocol and data processing pipelines are generally not disclosed. Given the increasing interest and demand of wearable sensors also for medical and diagnostic purposes, the metrological performance of such devices should be always considered, to be able to adequately interpret the results and always deliver them associated with the related measurement accuracy.•The sensor metrological performance should be always properly considered.•There are no standard methods for wearable sensors metrological characterization.•It is important to define rigorous test protocols, easily tunable for specific target applications.

Keywords: Measurement accuracy; Metrological characterization; Test protocol; Test protocol for the validation and the metrological characterization of wearable sensors for physiological monitoring; Validation; Wearable devices.

Graphical abstract

Fig. 1

Wearable sensors growth between 2015 and 2026: forecasts in terms of revenue (blue trend line) and sold units (green trend line) .

Fig. 2

Test setup, with test (Samsung Galaxy Watch3) and reference (Zephyr BioHarness 3.0) devices worn by the subject.

Fig. 3

Test protocol: acquisition phases.

Fig. 4

Distribution of measurement deviations (study case).

Fig. 5

Data agreement, Bland-Altman plot (case study).

Fig. 6

Correlation between test (y-axis) and reference (x-axis) data.