Mathematical models, smart shield for COVID-19 developed
A team of researchers has developed machine learning and mathematical models to study the severity of coronavirus cases in Oman. The team has also devised a smart shield to help screen temperature from a distance and help the wearer keep a safe distance.
The project, titled ‘Prediction, analysis and smart mask-based social distancing system for coronavirus disease (COVID-19),’ is headed by principal investigator Dr Abraham Varghese and co-principal investigator Dr Huda Salim al Shuaily from the University of Technology and Applied Sciences (Higher College of Technology).
It is among several other pioneering research projects of the COVID-19 Research Programme funded by the Ministry of Higher Education, Research and Innovation that has developed tools to tackle the COVID-19 situation in Oman. The smart shield helps by detecting body temperature from a distance.
‘When someone crosses the safe social distancing limit of 1m, the device indicates this through an LED signal. Once the safe distance is breached, the device will read the temperature of the person as well. If the temperature is more than 37.50°C, the device will sound an alarm. This will help in reducing risk of contracting the disease. The smart shield is very useful for health workers, police officers and other first responders who have to be in frequent contact with the public,’ a press release stated.
Dr Varghese and his team have also developed machine learning and mathematical models to study the severity of the coronavirus and analyse cases in Oman. The mathematical model calculates parameters based on various clinical stages and severity from available data in Oman. The team has come up with a machine-learning model to validate a clinical score that will identify potential of admission in an intensive care unit facility for a laboratory-confirmed SARS-CoV-2 case.
According to Dr Varghese, the team formulated a deterministic compartmental model (SEAMHCRD) that factors in various stages of infection (mild, moderate, severe and critical) to study the spread of COVID-19 and estimated the model parameters based on data available of the ongoing pandemic in Oman. It consists of nine nonlinear differential equations that take into account individuals who are susceptible, exposed, asymptomatic, hospitalised, critical and recovered patients. The summary of the findings is under consideration of publication in Scientific Reports by Springer Nature.
The research team has developed a web app (a CoV19-OM ICU calculator) with the machine-learning model to predict a clinical score that will identify patients who are likely to be admitted in an intensive care unit facility. It can facilitate better treatment management and optimise resource management.
Besides Dr Varghese and Dr Huda, the research team includes Dr Saad Salman Ahmed, Dr Eduardo M Lacap Jr, Dr Vinu Sherimon, Shajidmon Kolamban, Jagath Prasad Sreedhar, Syed Rafeek Ahmed, Dr Faryal Khamis, Dr Hasina al Harthi, Dr Hamed al Naamani, Dr Maher al Bahrani and Sunil Jacob.