OSCILLOMETRIC BLOOD PRESSURE ESTIMATION METHOD BASED ON MACHINE LEARNING

摘要

Provided by the present invention is an oscillometric blood pressure estimating method capable of providing accurate blood pressure estimating values and improved confidence interval (CIs). The oscillometric blood pressure estimating method based on machine learning according to an embodiment of the present invention comprises: (a) a step of obtaining the envelope of an oscillometric waveform about a measurement value based on a sample blood pressure measurement value; (b) a step of obtaining a physician maximum amplitude (PMA) based on the maximum amplitude (MA) of the envelope of the oscillometric waveform about the measurement value by using a non-paramilitary boot strap; (c) a step of obtaining a physician envelope (PE) based on the envelope of the oscillometric waveform about the measurement value by using the non-paramilitary boot strap; (d) a step of connecting the PMA with the PE; (e) a step of estimating a systolic blood pressure characteristic ratio (SBPR) and a diastolic blood pressure characteristic ratio (DBPR) about individual subject by using machine learning; and (f) a step of estimating the systolic blood pressure (SBP) and the diastolic blood pressure (DBP) with a confidence interval (CI) based on the estimated SBPR and the estimated DBPR.