Cuffless blood pressure estimation by error-correcting output coding method based on an aggregation of AdaBoost with a photoplethysmograph sensor

Abstract

This paper presented a novel cuffless and non-invasive technique of Blood Pressure (BP) estimation with a pattern recognition method by using a Photoplethysmograph (PPG) sensor instead of a cuff. Error-Correcting Output Coding (ECOC) method was adopted as a multi-classifier machine based on an aggregation of general binary classifiers. AdaBoost was applied as binary classifier machine. 368 volunteers participated in the experiment. The estimated Systolic Blood Pressure (SBP) was calculated from their individual information and several features of their Pulse Wave (PW). As a result of the comparison between measured SBP, estimated SBP, MD = -1.2 [mmHg] and SD = 11.7 [mmHg] were obtained. Hence, this technique would be helpful to the advance the development of continuous BP monitoring system, since the only one device to monitor BP is smaller than traditional measurements.