Key result
Smartphone camera algorithm estimates systolic blood pressure with ~14 mmHg deviation from standard cuffs.
Why the study?
Along with the rapidly developing market of apps and wearables introducing alternative blood pressure measurement methods, critical evaluation of these novel tools is needed.
Does a smartphone camera PPG algorithm reliably estimate systolic blood pressure compared to standard oscillometric measurement in hospital subjects?
Observational (n=1,000)
Single-blind
No
Does a smartphone camera PPG algorithm reliably estimate systolic blood pressure compared to standard oscillometric measurement in hospital subjects?
A novel smartphone camera PPG algorithm for estimating systolic blood pressure showed a mean absolute deviation of 14 mmHg compared to standard oscillometric measurement.
Answers are grounded in this paper and cite their sources.
Large error (14 mmHg MAD) indicates smartphone PPG BP estimates are not ready for clinical use; leaves open whether refined algorithms merit prospective validation.
Background: Along with the rapid developing market of novel apps and wearable devices several attempts to introduce alternative methods to measure blood pressure have evolved. Critical evaluation of these novel tools is needed. We tested a novel algorithm regarding its reliability and suitability to estimate systolic blood pressure using the photoplethysmographic (PPG) signal of a smartphone camera. Methods: Subjects at our University Hospital received a standard oscillometric cuff device (OCD) on their upper arm and a smartphone with its camera facing their index finger of the same arm after they signed informed consent. Seven sequential measurements starting with the OCD followed by the smartphone were performed. PPG files were analysed blinded for the oscillometric results and correlated with the mean between the preceding and the succeeding oscillometric result. In addition, 200 predefined subjects had repetitive measurement blocks within 3 days. Results: In total 1000 subjects were recruited and 4671 datasets were analysed. Mean absolute deviation was 14mmHg, standard deviation 11mmHg. Data is best displayed in fig. 1 as Bland-Altman plot. Subgroup analysis was performed for normotensive, grade 1 hypertensive and >grade 1 hypertensive individuals (table 1).
No takes yet. Share an insight, caveat, or question.
Eckstein et al. (2017) conducted an observational in Blood pressure measurement (n=1,000). Smartphone camera photoplethysmographic (PPG) signal algorithm vs. Standard oscillometric cuff device (OCD) was evaluated on Systolic blood pressure estimation accuracy (Mean absolute deviation 14 mmHg). A novel smartphone camera algorithm estimating systolic blood pressure had a mean absolute deviation of 14 mmHg (SD 11 mmHg) compared to standard oscillometric cuff measurements.
Synapse has enriched 10 related papers. Consider them for comparative context: