Assessment of respiratory disorders using MFCC and LPC applied to machine learning algorithms

Authors

  • Department of Electronics & Telecommunication, SSTC, Bhilai, Chhattisgarh, India
  • Department of Electronics & Telecommunication, SSTC, Bhilai, Chhattisgarh, India
  • Department of Biomedical Engineering, National Institute of Technology, Raipur, Chhattisgarh, India
  • OP Jindal University

Abstract

Respiratory Disorder detection using speech analysis is a vital research topic in today’s scenario, which is reliable, easy to use, economic, and efficient, and also helps in the detection of disorder in the early stage. The purpose of this article is to investigate whether or not speech parameters obtained can be utilized for envisaging respiratory disorders using predictive machine learning techniques, and to explore the role of different machine learning components such as data division protocols and classification to determine suitable speech parameters for detection of respiratory disorders. In this work, extraction and evaluation of speech parameters were done using PRAAT software. A dataset consisting of speech recordings and spirometry data was used in the experiment. This article used Mel Frequency Cepstral Coefficient (MFCC) and Linear Predictive Coding (LPC) coefficients for speech analysis. Two machine learning models, Support vector machine and Naïve Bayes have been employed and compared for the assessment of respiratory disorders. The results showed that SVM achieved the highest classification accuracy of 100% for LPC and 83.3% for MFCC utilizing the holdout method. While Naïve Bayes achieved the highest accuracy of 85% for LPC and 70% for MFCC using the 10-fold method. The results obtained indicated that the MFCC and LPC with high efficiency may provide an aid in the simple assessment of respiratory disorders with proposed classifiers.

Author Biographies

  • , Department of Electronics & Telecommunication, SSTC, Bhilai, Chhattisgarh, India

    Poonam Shrivastava is presently a Ph.D. research scholar at CSVTU, Bhilai (C.G) India. She received her B.E. degree in Electronics and Telecommunication engineering from Pt. Ravishankar Shukla University, Raipur (C.G) India in 2004 and MTech degree in Instrument and control from CSVTU, Bhilai in 2008. She has total teaching and research experience of more than 6 years. Her research interests include Signal Processing and machine learning. She has various research papers published in National and International Journals.

  • , Department of Electronics & Telecommunication, SSTC, Bhilai, Chhattisgarh, India

    Dr (Mrs.) Neeta Tripathi, a young and dynamic Principal, is the first woman and youngest Principal of the Engineering Institution recognized by the C.S.V.T.U. Bhilai in state of Chhattisgarh. She has an excellent track of academic record. She graduated in Electronics and Telecommunication Engineering from Government Engineering College Jabalpur in 1986 and obtained an M. E. degree in Electronics and Telecommunication with a specialization in Communication Systems in 1992 from the same Institute. Pandit Ravishankar Shukla University Raipur has awarded her Ph. D. Degree in Electronics and Telecommunication in 2007 on “Study of Face and Speech Parameters and Identification of their relationship for Emotional status Recognition”.   She had started her career as Research and Development Engineer in NITEL in 1987 and worked there for five years. In 1992, she had switched over from Industries to Teaching and serving this noble profession for more than twenty-nine years. During this tenure she worked in various capacities e.g. Lecturer, Reader, Professor, and Principal in various AICTE approved Institutes. She is actively engaged in Research Activities and published more than a hundred Research Papers in various International, National Journals/Conferences. She has received two best paper awards.  She has organized seven conferences/workshops and delivered a number of Keynote addresses and expert lecturers.  She has guided 6 numbers of Ph.D. thesis and a number of M. E. Dissertations. Presently she is guiding two Ph.D. candidates.  Her areas of research are Emotion Recognition, Signals Processing, Mobile Computing, Communication Systems, and Image Processing. She is a Member of IEEE, the Institution of Engineers (India), and MISTE. She is also a member of various Board of studies of CSVTU Bhilai. She is a recognized subject expert at CSVTU for the recruitment of faculty members

  • , Department of Biomedical Engineering, National Institute of Technology, Raipur, Chhattisgarh, India

    Dr. Bikesh Kumar Singh, Ph.D., received his doctorate degree in Biomedical Engineering from the National Institute of Technology, Raipur, India. He is currently working as an assistant professor in the Department of Biomedical Engineering of the National Institute of Technology, Raipur, India. He has more than11 years of teaching and research experience. He has published more than 60 research papers in various international journals and conferences of repute. He is a senior member of IEEE, a life member of IE, IETE, and CSI India. He has received IETE Gowri Memorial Award 2016, IEI Young Engineer Award 2012, Chhattisgarh Young ScientistAward-2010.His research interests include medical image processing and analysis, biomedical signal processing, soft computing, and machine learning

Downloads

Published

2024-02-16

Issue

Vol. 31 No. 1 (2024): Nonlinear Studies (NS)

Section

Articles

How to Cite

Assessment of respiratory disorders using MFCC and LPC applied to machine learning algorithms. (2024). Nonlinear Studies, 31(1). https://nonlinearstudies.com/index.php/nonlinear/article/view/3073