Detection of Covid-19 in a person through their Coughing and Breathing Patterns
Author(s):
Karthik Reddy G S, Darshan K M, Meghana G B, Nishita SIngh, Dr. Malatesh S H
Keywords:
Abstract
The new Coronavirus illness (Covid-19), according to the World Health Organization, has been a pandemic since March 2020. It starts out as a viral infection that causes swelling in the respiratory tract and can develop into a normal pneumonia. In fact, specialists emphasise the value of early COVID-19 virus infection diagnosis. By isolating the sick patients from others, the virus can be stopped from spreading. However, prompt assessment of breathing patterns is important for many medical emergencies. In this study, we provide a COVID-19 cough and breath analysis based on deep learning that can distinguish between positive COVID-19 cases and both negative and healthy COVID-19 cough, and breath captured on smartphones or wearable sensors. First, Mel Frequency Cepstral Coefficients (MFCC) will be used to reduce noise from audio signals, including cough and breath. After that, deep Long Term Short Memory (LSTM) model will be used to extract deep features. The proposed strategy produced the highest accuracy, over more than 80%, compared to the others in which the LSTM is utilized as a single model without any combination, according to performance data.
Article Details
Unique Paper ID: 156197
Publication Volume & Issue: Volume 9, Issue 3
Page(s): 426 - 430
Article Preview & Download
Share This Article
Join our RMS
Conference Alert
NCSEM 2024
National Conference on Sustainable Engineering and Management - 2024