Estimation of breathing signal and the respiratory parameters from the speech recordings using honey badger-based modular neural system

• Published in: Multimedia tools and applications 2024-02, Vol.83 (30), p.73957-73982
• Main Authors: Patil, Mangal, Naregalkar, Prajakta R, Tondare, Sharda, Morbale, Jyoti, Patil, Aparna S.
• Format: Article
• Language: English
• Subjects: Badgers; Breathing; Chronic obstructive pulmonary disease; Computer Communication Networks; Computer Science; Data collection; Data Structures and Information Theory; Medical research; Modular systems; Multimedia Information Systems; Neural networks; Noise prediction; Parameter estimation; Parameter identification; Parameter robustness; Respiration; Signal quality; Special Purpose and Application-Based Systems; Speech; Structural health monitoring; Telemedicine
• ISSN: 1573-7721; 1380-7501; 1573-7721
• DOI: 10.1007/s11042-024-18353-2

Breathing is a necessary mechanism for speech production. Estimating breathing signal and respiratory parameters are the most important study in medical research. The breathing signal can be evaluated using a variety of models. However, the relevant result was not attained because of the poor quality of the model. Breathing is a vital bodily function, and changes in breathing patterns can indicate underlying health problems. For example, people with respiratory conditions such as asthma, Chronic Obstructive Pulmonary Disease (COPD), and heart failure may have abnormal breathing patterns. The speech data contains more noise content, which maximizes the model’s complexity. Analyzing the speech recordings makes it difficult to develop a non-invasive and continuous monitoring system for respiration because the system failed to estimate the breathing signal using traditional methods. These problems resulted in poor prediction and less accuracy for prediction. So, this research introduces a novel Honey Badger-based Modular Neural System (HBMNS) to the challenging task of using voice recordings to estimate respiratory parameters and breathing signals. This study aims to bridge the gap between speech analysis and respiratory health monitoring by leveraging the adaptability and robustness observed in honey badger behavior. This indicates an interest in identifying patterns or features in speech that correlate with respiratory behavior. In addition to breathing signals, the system aims to estimate various respiratory parameters. These parameters might include breath rate, tidal volume, and breath event. Initially, data collection and importation into the Python environment involved speech recording data. The data was then pre-processed, and honey-badger optimization was used to extract the spectral features. The breathing signal and respiratory parameters are then estimated by a modular neural network. The provided model achieved a good exactness score in the estimation after the performance was measured.