Rapid Computer Diagnosis for the Deadly Zoonotic COVID-19 Infection

Abstract

The life cycle of SARS-CoV-2 is complexly linked with that of its host, thereby, rendering all prospective treatments ineffective. Recently, there was a drift from Cross-species transmission (Zoonosis) → Intra-species → Nosocomial transmission, thereby, increasing the risk of infection. In consortium with WHO, rapid computer diagnosis (RCD) was exigent, as it will increase the chances of identification of suspected cases and minimize false-positive diagnosis. Etaware-CDT-2020 RCD Model “Y = α + β1X1 + β2X2 + β3X3 + … β26X26” was developed using broad-spectra symptoms catalogue for COVID-19. The best-fit model was adjudged by R2, R-SqAdj, AIC, BIC, MSEPred., MAE, LOO_Press, LOOPreR2, LOO-MAE, LGO_Press, LGOPreR2, LGO-MAE etc., validated by bootstrapping and trial diagnosis. The R2 and R-SqAdj values were positive (1.00 and 1.00, respectively), while AIC and BIC values were negligible (−3677.10 and −3659.60, respectively). The mean error of diagnosis was least in Hubei cases (11.1), while the standard error of diagnosis was insignificant in confirmed cases outside Hubei (2.0), and those linked (or not) to Wuhan (2.0). The similarity index of diagnosis (R and R2) was best-fit in Hubei cases (0.78 and 0.49, respectively). Etaware-CDT-2020 is a better alternative for COVID-19 diagnosis and it is very easy to setup. It can be utilized in hospitals, clinics, homes, offices, and public places with ease.