Sensor fusion in human blood pressure measurements using oscillometric and auscultatory principles: Error analysis and implementation. (January 2022)

• Record Type: Journal Article
• Authors: Tavares, Noel; Vetrekar, Narayan; Gad, R.S.; Kakodkar, Uday
• Is Part Of: Biomedical signal processing and control. Volume 71(2022)Part A
• Publication Date: 2022-01
• Subjects: Human blood pressure -- Systolic -- Diastolic -- Oscillometric -- Auscultatory -- 'Korotkoff' sounds -- Windkessel model; 00-01 -- 99-00; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2021.103081
• Languages: English
• ISSNs: 1746-8094

Abstract

Highlights: Complementary information fusion of auscultation and oscillometric signals. Measure real-time pressure using auscultation and oscillometric principle. Windkessel model to simulate human heart. Extensive benchmark results based on the optimized fusion scheme. Abstract: The accurate measurement of human blood pressure is a growing needs in the recent years, to obtain the prior information about the heart related risks. Majority of the available blood pressure measurement systems are based on the principle of auscultation and oscillometric method, operated independently. In this work, we develop a low-cost blood pressure measurement system based on auscultation and oscillometric principle. We present the complementary information fusion approach to combine auscultation and oscillometric signals to improve overall accuracy of the system. However, the parameters for our complementary information fusion is optimized based on the simulated sample data points generated using Windkessel model. We presents the extensive experimental validation results to measure systolic and diastolic pressure for three different categories of patients that includes Hypotension, Normal and Hypertension. The better average measurement accuracy of 90.84 % is obtained using our approach of complementary information fusion for α = 0.5 value on simulated sample data. Further, we use the optimized complementary fusion operator with α = 0.5 in our developed system and the measurement obtained with our method is compared against the reference meter with minimum measurement errors, demonstrating the potential of our system in measuring accurately the blood pressure.