Energy-Efficient Respiratory Sounds Sensing for Personal Mobile Asthma Monitoring

Abstract

Current medical practice of long-term chronic respiratory diseases treatment lacks a convenient method of empowering the patients and caregivers to continuously quantitatively track the intensity of respiratory symptoms. Such is "asthmatic wheezing", occurring in respiratory sounds. We envision a mobile, personalized asthma monitoring system comprising of a wearable, energy-constrained acoustic sensor and smartphone. In this article we address the energy-burden of acquisition and streaming of acoustic respiratory signal, and lessen it by applying the concept of compressed sensing (CS). First we analyse the adherence of normal and pathologic respiratory sounds frequency representations (DFT, DCT) to the sparse signal model. Given the pseudo-random non-uniform subsampling encoder implemented on MSP430 microcontroller, we review tradeoffs of accuracy and execution time of different CS algorithms, suitable for realtime respiratory spectrum recovery on smartphone. Working CS respiratory spectrum acquisition prototype is demonstrated, and evaluated. Prototype enables for real-time reconstruction of spectra dominated by approximately 8 frequency components with more than 80% accuracy, on Android smartphone using OMP algorithm, from only 25% signal samples (w.r.t. Nyquist rate) acquired and streamed by sensor at 8 kbit/s. Index Terms-M-health, asthmatic wheezing, compressive sensing, non-uniform sampling, orthogonal matching pursuit.1530-437X (c)

System-Level Power Consumption Analysis of the Wearable Asthmatic Wheeze Quantification

“…Here, we evaluate power requirements for the CS signal encoding implemented by sub-Nyquist sampling of the analog input signal at the nonuniform pseudorandomly spaced sampling instants [30,31]. The choice of encoder was motivated by the fact that the mentioned CS encoder design requires minimal number of digitized signal samples, thus enabling the highest savings of active power in signal acquisition subsystem (i.e., ADC's and MCU's active time spent on acquisition, data handling, and storage) [71].…”

“…Also, we provide generalized guidelines on hardware component architectures best fitting the application. The analysis builds up upon our extensive prior research on novel energy-efficient signal acquisition and wireless transport schemes [30], design of specialized low-power wheeze recognition algorithms suitable for running onboard energy-constrained devices (sensor node and smartphone) [12,23], and verification of all subsystems on several hardware laboratory prototypes [12,22,31,32].…”

“…In order to lower the data rate and consequently the average power required for streaming, the signal from the sensor to a smartphone, the second scenario, exploits the compressibility of respiratory sounds in the frequency domain [30,33] and applies a concept of compressed sensing (CS) [34,35]. By combining signal acquisition and compression in a single linear transformation step (i.e., CS encoding), CS in comparison to classic audio coding techniques [36,37] enables simultaneous lowering of the data rate, while retaining the low complexity of signal processing on the sensor.…”

“…The compressed signal is streamed over to the smartphone. By knowing the subsampling pattern, the CS decoder subsystem block obtains a reconstruction of the compressible STFT spectra from the CS-encoded signal [30,32]. Finally, a robust hidden Markov model-(HMM-) based frequency line tracking algorithm [23] implemented on a smartphone [22] enables for wheeze detection from reconstructed spectra at less than 5% loss of classification accuracy.…”

Design Analysis and Human Tests of Foil-Based Wheezing Monitoring System for Asthma Detection

“…While the frequency of a wheeze from the trachea lies in the range of 100-2500 Hz, it is reduced to 100-1000 Hz from the chest because lung tissue, chest wall, and skin absorb the higher frequencies before they reach our sensor [5,6]. The chest-wall tissue acts as a low-pass spectral constraint on respiratory sounds, which, when measured on the skin surface with an acoustic sensor (microphone or accelerometer) positioned on human chest, back or neck, typically reside in the frequency band below 1 kHz [26]. Excess amount of chest hair may cause interference with the adhesion and performance of the sensor.…”