Real time detection and forecasting technique for asthma disease using speech signal and DENN classifier. (July 2022)

• Record Type: Journal Article
• Authors: Iqbal, MD. Asim; Devarajan, Krishnamoorthy; Ahmed, Syed Musthak
• Is Part Of: Biomedical signal processing and control. Volume 76(2022)
• Publication Date: 2022-07
• Subjects: Asthma detection -- Differential evolutionary neural network (DENN) -- IWO algorithm -- Respiratory sounds -- Real time detection and forecasting; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2022.103637
• Languages: English
• ISSNs: 1746-8094

Abstract

Highlights: Asthma, this is an abnormal lung sound associated with a respiratory disorder. It shows the sinusoidal continuum features in the time domain and an important characteristic of the system in the spectrogram. RTDF technique used to classify normal and asthmatic subjects based on the alterations in their voice signals. The implementation is done with MATLAB tool and analyzes the performance in terms of probability of correct classification. The presentation of the planned process has been evaluated using lung sound from patients and normal subject under special Signal-to-Noise Ratio (SNR) Abstract: Abnormal sound of the lungs associated with asthma and respiratory disease. On the spectrogram, it exhibits continuous sinusoidal qualities across time as well as significant computer properties. In this article, a voice signal and an optimum classifier is used to present a real-time detection and forecasting (RTDF) approach for Asthma illness. For asthma diagnosis and forecasting, the suggested RTDF approach employs the improved whale optimization (IWO) algorithm. RTDF technology is used to classify normal and asthmatic diseases based on changes in voice signals. RTDF technology integrates a variety of working Differential evolutionary neural network (DENN) classifiers that are superior to current SVM classifiers and help prevent the development of asthma suppression. In addition, the digital detection gateway-based secondary output can detect or reject the primary output, making breath detection more reliable for weak respiratory sounds. Implementation is done in conjunction with MATLAB tools and performs performance analysis based on the probability of correct classification. The presentation of the planned process was evaluated using a special signal-to-noise ratio (SNR) using lung sounds from patients and normal subjects.