Acoustic-pressure sensor array system for cardiac-sound acquisition. (August 2021)

• Record Type: Journal Article
• Authors: Wang, Tian; Gong, Meihui; Yu, Xiaoyu; Lan, Guangdong; Shi, Yunbo
• Is Part Of: Biomedical signal processing and control. Volume 69(2021)
• Publication Date: 2021-08
• Subjects: ECM sensor -- Sensor array -- Cardiac-sound acquisition -- Data fusion; RMS Root mean square -- ADC Analog-digital converter -- dBFS Decibel full scale -- ECM Electret-condenser-microphone -- FIR Finite impulse response -- FFT Fast Fourier Transform -- FPGA Field-programmable gate array -- MCU Microcontroller unit -- OSR Oversampling rate -- PVDF Polyvinylidene fluoride -- SNR Signal-to-noise ratio; 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.2021.102836
• Languages: English
• ISSNs: 1746-8094

Abstract

Highlights: A cardiac-sound-detection system with acoustic-pressure sensor array is proposed. Field-programmable gate array and microcontroller are used for signal processing. The system shows reduced noise and comparable performance to conventional system. Low cost and thin structure allow integration with wearable electronics. Abstract: This paper presents a cardiac-sound-detection system based on an acoustic-pressure sensor array. The traditional cardiac-sound probe has the disadvantages of high cost and thick structure, which makes embedding difficult. Therefore, a methodology for cardiac-sound detection, comprising a sound-pressure sensor array, is proposed to achieve a low cost, very thin, and wearable cardiac-sound probe, which can be embedded in furniture or clothes for long-term heart sound monitoring. By introducing several miniature electret-condenser-microphone-based acoustic-pressure sensors, a sensor array for picking up cardiac sound was designed. The data-processing platform comprised a field-programmable gate array and microcontroller unit. Furthermore, a series of de-noising methods including oversampling, finite impulse response digital filtering, and digital extraction were used, thereby facilitating accurate cardiac-sound signal identification. Test results showed that the average Root Mean Square amplitude of the noise after processing was more than 3 dB lower than that of the signal obtained from a single sensor in the array. The signal-to-noise ratio improved more than 7 dB after processing. Furthermore, actual cardiac sounds were acquired and compared using the proposed sensor array and the dedicated cardiac-sound sensor. The proposed 4 mm thick sensor array could effectively acquire S1- and S2-segments of cardiac sound. The proposed system demonstrates the potential for use in the development of wearable cardiac-sound monitoring systems.