Improving the Response rate of RT PCR testing using deep convolutional encoder

Abstract

The sensitivity of the RT-PCR test is limited, and the method itself is laborious and time-consuming to carry out. When it comes to making a diagnosis of COVID-19, chest CT scans have the potential to be of assistance. If, on the other hand, the infection does not spread to the lungs, then an abnormality check with a CT scan will not be necessary. Complementary assays have the potential to considerably reduce the frequency of false-positive results when they are used in combining with RT-PCR or CT scans. We propose in this study a decision support system that is based on deep learning and has the capability to evaluate the practically commonly required laboratory parameters for the purpose of identifying COVID-19. Here dataset contains the 1428 radiographs (224 COVID-19, 504 Healthy, 700 Pneumonia) are considered for experimentation.

Keywords

• RT PCR
• Response rate
• Deep convolutional encoder
• Testing