Pervasive computational model and wearable devices for prediction of respiratory symptoms in progression of COVID-19

The purpose of this paper is to provide a model for prediction of respiratory symptoms in the progression of COVID-19, social distancing, frequent hand washes, wearing of face mask in public are some of the potential measures of preventing the disease from further spreading. In spite of the effects and efforts taken by governments, the pandemic is still uncontrolled in major cities of the world. The proposed technique in this paper introduces a non-intrusive and major screening of vital symptoms and changes in the respiratory organs.

The novel coronavirus or Covid-19 has become a serious threat to social and economic growth of many nations worldwide. The pace of progression was significantly higher in the past two months. Identified by severe respiratory illness, fever and coughs, the disease has been threatening the lives of human society. Early detection and prognosis is absolutely necessary to isolate the potential spreaders of the disease and to control the rate of progression.

Recent studies have highlighted the changes observed in breathing characteristics of infected patients. Respiratory pattern of Covid-19 patients can be differentiated from the respiratory pattern of normal cold/flu affected patients. Tachypnoea is one among the vital signs identified to be distinguishing feature of Covid-19. The proposed respiratory data capture will commence with facial recognition, use of infrared sensors and machine-learning approaches to classify the respiratory patterns, which finally narrows down as a symptom of Covid-19.

Proposed system produced outcome of 94% accuracy, precision, recall and a F1-measure as an average in the conducted experiments. This method also proves to be a fruitful solution for large-scale monitoring and categorisation of people based on the symptoms.