Artificial Intelligence Approach to Reduce PCR Tests

Nowadays, the Covid-19 crisis is the biggest global challenge to control. All scientists have been working during the last two years to find solutions for this crisis in numerous ways. Therefore, identifying a person having Covid-19, before entering public places and mainly to the hospitals is a timely need to reduce the transmission of the virus. Polymerase chain reaction (PCR) test is the primary diagnostic tool used to identify Covid-19 patients. However, due to the vast number of patients, the demand for PCR diagnostic assays cannot be met. This research study was primarily concerned with eliminating PCR testing to a certain extent through intelligent approaches. This paper discusses a model using computer vision-based approaches for detecting Covid-19. The automated testing technique, which is based on a questionnaire and eye color scanning, was performed with artificial intelligence-based image processing. This device is capable of recognizing the Covid-19 patients before they enter into hospitals or public locations. The CNN model was developed using an open-source data set provided by the world health organization (WHO). The result demonstrates an 89% accuracy, concluding that this system indicates an excellent prediction performance for the Covid-19 diagnosis.