EVALUATION OF A PROTOTYPE CUFFLESS MULTI-SENSOR DEVICE TO TRACK BLOOD PRESSURE CHANGES

Christine Hove,Frode Sæter,Alexey Stepanov,Kasper Gade B⊘tke-rasmussen,Trine M. Seeberg,Espen Westgaard,Sondre Heimark,Bård Waldum-grevbo,Jonny Hisdal,Anne Cecilie K. Larstorp
DOI: https://doi.org/10.1097/01.hjh.0001020064.58532.24
IF: 4.9
2024-05-01
Journal of Hypertension
Abstract:Objective: Accurate measurements are essential if cuffless blood pressure (BP) devices are to be used in hypertension care. The aim of the present study was to evaluate the ability of a prototype cuffless device (XXXXX) to track changes in BP compared to a non-invasive, continuous BP monitor (reference BP; Human NIBP Nano system, ADInstruments, Dunedin, New Zealand) in an experimental laboratory set up. Design and method: After ethical approval, three external techniques (Figure 1) to induce BP changes via different physiological mechanisms, were applied in 23 healthy subjects. Continuous data from the prototype cuffless device on the upper arm was obtained simultaneously with reference BP measurements. The cuffless device measures skin temperature, 3D acceleration, 3D angular velocity and photoplethysmography. Data from 16 of the participants were used to train a machine learning model, which was then used to estimate BP (15-second resolution) on data from 7 other participants. The cuffless BP device was calibrated once using brachial cuff BP measurement (WatchBP O3, Microlife, Cambridge, United Kingdom). The present study provides preliminary results from a larger on-going study designed to evaluate performance of the cuffless device according to ISO 81060-3:2022 standard. Results: A timeseries plot displaying BP measurements from the cuffless and the reference device is displayed in Figure 2. The mean error (± corrected standard deviation of error) was 2.4 (± 7.9) mmHg for systolic blood pressure (SBP) and 1.7 (± 7.4) mmHg for diastolic blood pressure (DBP) for the accuracy requirement. The corresponding errors for stability requirements were 1.9 (± 8.1) and 1.9 (± 6.0) mmHg for SBP and DBP. Conclusions: The present study demonstrated that the prototype cuffless BP device could track BP changes with high accuracy in healthy subjects in a laboratory setting. The cuffless device fulfilled the acceptance criteria (± 6 (± 10) mmHg) for the accuracy and stability requirements from the ISO 81060-3:2022 standard. Further research is necessary to determine accuracy of the prototype cuffless BP device during real-life conditions as described in the 2023 ESH recommendations for the validation of cuffless devices.
peripheral vascular disease
What problem does this paper attempt to address?