Detection and Classification of Abnormal Respiratory Sounds on a Resource-constraint Mobile Device

Abstract

Abnormal breath sounds like wheezes, crackles and stridor at times manifest similar morphologies and pathological features of lung airways obstruction. This may pose problems to proper diagnosis and evaluation of the underlying respiratory condition by human auscultation. In this study, the authors experimented with Time-Frequency threshold-dependent (TFTD) algorithm for detection and classification of breath sounds based on Smartphone. The TFTD algorithm computes important and distinct features of each breath sound using spectro-temporal analysis of recorded lung sounds which can enhance qualitative measurement and quantitative indexing of different respiratory sounds. Several algorithms which run exclusively on desktop computers have been developed for detecting and analyzing specific lung sounds such as wheezes. However, few attempts have been made to perform such analysis on portable devices like mobile phones due to computational complexities and high power consumption associated with the analyses. Our experimental results demonstrate that recent smartphones with improved computational capacity are able to provide comparative performance on analysis of respiratory signals. Furthermore, these phones can serve as convenient tools for measuring and detecting early signs of pulmonary disorders particularly at home and during ambulatory care services where conventional and specialized medical devices may not be accessible.

Design of Wearable Breathing Sound Monitoring System for Real-Time Wheeze Detection

“…Thus, Refs. [ 18 , 38 , 41 ] and our proposed system have lower computational complexities than Refs. [ 20 , 39 , 40 ].…”

“…Moreover, the quartile frequencies in the power spectrum of asthma group were higher than that of healthy control group, and they considered that wheezing sounds contained the peak frequency which was above 150 Hz and was at least three times higher than the baseline level. Uwaoma et al developed a time frequency threshold-dependent (TFTD) algorithm to detect wheezing sounds in a smartphone [ 38 ]. The peak frequency, the number of consecutive harmonics, and the duration was used as the wheezing features, but they also indicated that the peak frequency of breathing sounds, that is highly similar to that of wheezing sounds, would easily result in the detecting failure.…”

“…Refs. [ 18 , 38 , 41 ] and our proposed system processed spectra from fast Fourier transform (FFT) to obtain the features of wheezes, and their computational complexities are all in . Thus, Refs.…”

“…In Refs. [ 18 , 38 , 41 ], they have to use classifiers to recognize wheezes, but our proposed system only uses addition and compares with thresholds, so our proposed system has lower computational complexity than Refs. [ 18 , 38 , 41 ].…”

Automatic Wheezing Detection Based on Signal Processing of Spectrogram and Back‐Propagation Neural Network

“…Thus, while these approaches enable precise wheezing detection, they are slow. Recent attempts to achieve higher sensitivity and efficient detection performance include the consideration of a set of criteria in the time-frequency domain [15][16][17][18][19][20][21][22][23]. These criteria pertain to the duration, pitch range, and magnitude of wheezes in the time-frequency representation of the wheezes obtained through spectrogram analysis.…”