Blood pressure estimation using smartphone

—This paper presents an experimental cuff-less measurement of systolic (SBP) and diastolic blood pressure (DBP) using smartphone. A photoplethysmographic signal (PPG) measured by a smartphone camera is used to estimate blood pressure (BP). This paper contains comparison of several machine learning (ML) methods for BP estimation. Filtering the PPG signal with a band-pass filter (0.5-12 Hz) followed by feature extraction and using Random Forest (RF) methods separately or as a weak regressor in adaptive boosting (AdaBoost) or boot-strap aggregating (Boosting) reached the best results according to Association for the Advancement of Medical Instrumentation (AAMI) and British Hypertension Society (BHS) standards among all regression ML models. The mean absolute error (MAE) and standard deviation (SD) of Bagging model were 4.532 ± 3.760 mmHg for SBP and 2.738 ± 3.032 mmHg for DBP (AAMI). This result meets the criteria of the AAMI standard.