Pay Attention to the cough: Early Diagnosis of COVID-19 using Interpretable Symptoms Embeddings with Cough Sound Signal Processing

COVID-19 (coronavirus disease 2019) pandemic caused by SARS-CoV-2 has led to treacherous and devastating catastrophe for humanity. At the time of writing, specific antivirus drugs or vaccines are recommended to control infection and spread. The current diagnosis of COVID-19 is done by-Transcription Polymer Chain Reaction (RT-PCR) testing. However, this is expensive, time-consuming, and not easily available in straitened. An interpretable and COVID-19 diagnosis AI framework is devised and based on the cough sounds features and symptoms metadata to overcome limitations. The proposed framework's performance was evaluated using a dataset containing Symptoms and Demographic data of 30000 audio, 328 cough sounds from 150 patients with four cough classes (-19, Asthma, Bronchitis, and Healthy). Experiments' results show that the captures the better and robust feature embedding to distinguish between-19 patient coughs and several types of non-COVID-19 coughs with higher and accuracy of 95.04 \± 0.18% and 96.83\± 0.18%, all the while maintaining interpretability.