November 18, 2022

COVID-19 Detection Research by IITKGP Foundation Scholar Mahil Shaik Hussain

IITKGP Foundation 


For the past several years the IIT KGP Foundation USA has been supporting exemplary talents from the Insti to undertake research work in top global universities. This Summer, the Foundation facilitated Mahil Shaik Hussain to pursue his work at the University of Quebec on the importance of different cough phases for COVID-19 detection. 

 

Cough is an important symptom of numerous respiratory diseases, including COVID-19. While different cough phases (i.e., inhalation, compression, and expulsion) have been shown to be related to different pathological origins, existing cough-based COVID-19 detection systems rely on the entire cough recording, thus such phase-related characteristics are overlooked. 

 

In this study, the research team including Hussain annotated over 1,250 cough recordings from two publicly-available cough sound databases. They provided the research community with fine-grained cough phase labels. Furthermore, they extracted a number of temporal and acoustic features from each cough phase and test their usefulness and complementarity for COVID-19 detection. After extracting the features they looked at the crucial features in diagnosing the disease through experimentation and data analysis.

 

Their findings showed the importance of the coughing phase segmentation for improved COVID-19 detection. The findings were also found to be significant for the development of models that are interpretable and can better generalize across datasets. 

 

Two cough sound datasets were employed in this study, namely the ComParE and DiCOVA2 datasets. In both datasets, participants were asked to make forced coughs several times in a quiet environment. The cough sounds were collected remotely via smartphones or computers. Cough recordings were re-sampled to 16kHz in ComParE and 44.1kHz in DiCOVA2. 

 

Extracted features can be categorized into 2 categories, one being the “opensmile” features using the library available in python and one being the duration of different segments in the cough and results have been inclined over the positive impact of this duration as a crucial

feature over the opensmile features extracted.

 

Do you want to support young Kgpians like Shaik Mahil Hussain? Write to us at meenakshi@iitkgpfoundation.org