Elimination of artifacts in oscillometric waveforms using empirical wavelet transform to improve accuracy of blood pressure estimation

Blood pressure measurement is vital in healthcare and many methods have been developed for this purpose of which the oscillometric technique is the most popular. However, accurate estimation of BP using an oscillometric waveform remains a challenge, particularly when the waveform is distorted by noise and artifacts, such as due to involuntary motion. In response, this paper proposes a band pass filtering (BPF) method to deliver a filtered oscillometric signal that is suitable for blood pressure estimation using envelope detection. To validate the utility of the proposed method, two algorithms are implemented to estimate the systolic and diastolic BPs for various operating conditions. The obtained results indicate that a properly designed BPF exhibits significant improvement in delivering accurate BP measurement. It is envisioned that the proposed method is suitable for implementation in emergency medical devices, such as for blood pressure monitoring in an ambulance under significant motion artifacts.