EAP-DL: Enhanced asthma prediction with voice recording using efficient feature extraction and classification technique

Document Type : Primary Research paper


1 Ph.D Scholar, Department of ECE, Annamalai University, Tamil Nadu, INDIA

2 Assistant Professor, Department of ECE, Annamalai University, Tamil Nadu, INDIA

3 Professor, Department of ECE, SR University, Warangal, Telangana, INDIA


Asthma is determined flight courses disease depicted by irregular attacks of shortness of breath and wheezing. Adherence to tranquilize frameworks is an average crashing and burning for asthmatic patients and there exists a need to screen such patients' adherence. The ID of internal breaths from accounts of inhaler use can give test confirmation about patients' adherence to their asthma remedy framework.This paper proposes an enhanced asthma prediction with voice recording using deep learning (EAP-DL) classifiers. Firstly, the improved Ripplet-II Transform(IR2T) algorithm were worn to detect dissimilar type of breathing sound in loud signal coming from the container in the similar incidence variety as breathing. Secondly, recurrent deep neural network (RDNN) classifier used to analysis and predict the asthma disease at earlier. The dataset was gotten from not many asthma outpatients who went to a respiratory facility over a multi month time span. Assessment of the calculation on this dataset accomplished high affectability explicitness and precision of identifying inward breaths contrasted with manual inward breath identification. The presentation of the proposed strategy has been assessed utilizing lung sounds from patients and ordinary subjects under various Signal-to-Noise Ratio (SNR).