Wavelet-based embedded algorithm for respiratory rate estimation from PPG signal. (July 2017)

• Record Type: Journal Article
• Authors: Lin, Yue-Der; Chien, Ya-Hsueh; Chen, Yi-Sheng
• Is Part Of: Biomedical signal processing and control. Volume 36(2017)
• Page Start: 138
• Page End: 145
• Publication Date: 2017-07
• Subjects: Complex Morlet wavelet -- Photoplethysmography (PPG) -- Pulse oximeter -- Respiratory-induced intensity variation (RIIV) -- Respiratory rate (RR); 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.2017.03.009
• Languages: English
• ISSNs: 1746-8094

Abstract

Highlights: A wavelet-based algorithm is presented for the RR estimation from PPG signal. The estimation results agree well with those from usual spectrum analysis methods. The experiments of long-term trace, breath-holding and paced-breathing are conducted. The experimental results indicate that the proposed algorithm is reliable. The algorithm is feasible to be incorporated in the commercial pulse oximeter. Abstract: Photoplethysmography (PPG) is a popular technique utilized in pulse oximeter. Several researches have shown that PPG signal possesses the respiratory-induced intensity variation (RIIV) component, which implies that arterial oxygen saturation, heart rate (HR) and respiratory rate (RR) can be acquired by a single device. The commercial pulse oximeter generally provides the values of arterial oxygen saturation (SpO2 ) and HR. To successfully add the function of RR estimation to pulse oximeter, an algorithm requiring fewer resources plays a critical role. This paper presents a wavelet-based algorithm for RR estimation from PPG signal that can be implemented in the micro-controller (MCU) of pulse oximeter. The algorithm has been coded in C language and tested in a 32-bit MCU. The estimation results derived by the algorithm agree well with those from usual spectrum analysis methods. The RR estimations derived by PPG and respiratory signal are analyzed by Bland–Altman method. The RR estimations for long-term trace, breath-holding and paced-breathing experiments are also conducted to verify the performance of the proposed algorithm. The experimental results indicate that the proposed algorithm is highly reliable and is feasible to be incorporated in the commercial pulse oximeter.