Analysis of Acoustic Features for Speech Sound Based Classification of Asthmatic and Healthy Subjects

Non-speech sounds (cough, wheeze) are typically known to perform better than speech sounds for asthmatic and healthy subject classification. In this work, we use sustained phonations of speech sounds, namely, /α:/, /i:/, /u:/, /eI/, /o℧/, /s/, and /z/ from 47 asthmatic and 48 healthy controls. We consider INTERSPEECH 2013 Computational Paralinguistics Challenge baseline (ISCB) acoustic features for the classification task as they provide a rich set of characteristics of the speech sounds. Mel-frequency cepstral coefficients (MFCC) are used as the baseline features. The classification accuracy using ISCB improves over MFCC for all voiced speech sounds with the highest classification accuracy of 75.4% (18.28% better than baseline) for /o℧/. The exhale achieves the highest classification accuracy of 77.8% (4.2% better than baseline). Comparable accuracies using speech sound /o℧/ and non-speech exhale indicate the benefit of the rich acoustic features from ISCB. An analysis of 21 ISCB features groups using forward feature group selection shows that loudness and MFCC groups contribute the most in the case of /o℧/, with interquartile range between 2nd and 3rd quartile of loudness feature being the best discriminator feature.

[1]  E. Andrès,et al.  Analysis of Respiratory Sounds: State of the Art , 2008, Clinical medicine. Circulatory, respiratory and pulmonary medicine.

[2]  D. Andina,et al.  Feature selection using Sequential Forward Selection and classification applying Artificial Metaplasticity Neural Network , 2010, IECON 2010 - 36th Annual Conference on IEEE Industrial Electronics Society.

[3]  M. Bahoura,et al.  Respiratory sounds classification using Gaussian mixture models , 2004, Canadian Conference on Electrical and Computer Engineering 2004 (IEEE Cat. No.04CH37513).

[4]  Markus Neuhäuser,et al.  Wilcoxon Signed Rank Test , 2006 .

[5]  Hans Pasterkamp,et al.  Automated Spectral Characterization of Wheezing in Asthmatic Children , 1985, IEEE Transactions on Biomedical Engineering.

[6]  Florian Eyben,et al.  Real-time Speech and Music Classification by Large Audio Feature Space Extraction , 2015 .

[7]  Ismail Kocak,et al.  Subjective and objective evaluation of voice quality in patients with asthma. , 2007, Journal of voice : official journal of the Voice Foundation.

[8]  M. Jaeger,et al.  Mechanical aspects of coughing , 1975, Pneumonologie.

[9]  M. Oud,et al.  Lung function interpolation by means of neural-network-supported analysis of respiration sounds. , 2003, Medical engineering & physics.

[10]  Prasanta Kumar Ghosh,et al.  Comparison of Cough, Wheeze and Sustained Phonations for Automatic Classification Between Healthy Subjects and Asthmatic Patients , 2018, 2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC).

[11]  Fabio Valente,et al.  The INTERSPEECH 2013 computational paralinguistics challenge: social signals, conflict, emotion, autism , 2013, INTERSPEECH.

[12]  Sao-Jie Chen,et al.  Automatic Wheezing Detection Based on Signal Processing of Spectrogram and Back-Propagation Neural Network. , 2015, Journal of healthcare engineering.

[13]  Chih-Jen Lin,et al.  LIBSVM: A library for support vector machines , 2011, TIST.

[14]  Bradley W. DeVrieze,et al.  Peak Flow Rate Measurement , 2019 .

[15]  Barry M. G. Cheetham,et al.  DSP algorithm for cough identification and counting , 2002, 2002 IEEE International Conference on Acoustics, Speech, and Signal Processing.

[16]  A. Hamdan,et al.  Phonatory Symptoms and Acoustic Findings in Patients with Asthma: A Cross-Sectional Controlled Study , 2014, Indian Journal of Otolaryngology and Head & Neck Surgery.

[17]  Kenneth Sundaraj,et al.  A survey on automated wheeze detection systems for asthmatic patients. , 2012, Bosnian journal of basic medical sciences.

[18]  J Sundberg,et al.  Effects of lung volume on the glottal voice source. , 1998, Journal of voice : official journal of the Voice Foundation.

[19]  A. Rebuck,et al.  Lung volume changes in asthma measured concurrently by two methods. , 2015, The American review of respiratory disease.

[20]  F. Martinez,et al.  Genes, environments, development and asthma: a reappraisal , 2006, European Respiratory Journal.

[21]  R. Jane,et al.  Spectral analysis of respiratory sounds to assess bronchodilator effect in asthmatic patients , 1998, Proceedings of the 20th Annual International Conference of the IEEE Engineering in Medicine and Biology Society. Vol.20 Biomedical Engineering Towards the Year 2000 and Beyond (Cat. No.98CH36286).

[22]  G. Charbonneau,et al.  Digital processing techniques of breath sounds for objective assistance of asthma diagnosis , 1982, ICASSP.

[23]  Tomasz P. Zielinski,et al.  Joint Application of Audio Spectral Envelope and Tonality Index in an E-Asthma Monitoring System , 2015, IEEE Journal of Biomedical and Health Informatics.

[24]  D. Scuse,et al.  A comparison of neural network models for wheeze detection , 1995, IEEE WESCANEX 95. Communications, Power, and Computing. Conference Proceedings.

[25]  Kenneth Sundaraj,et al.  Identification of asthma severity levels through wheeze sound characterization and classification using integrated power features , 2019, Biomed. Signal Process. Control..