Submit Article
Back to Journal

SM Journal of Infectious Diseases

Classifier for Prediction of the Level of Disease COVID-19

[ ISSN : 3066-134X ]

Abstract
Details

Received: 06-Dec-2022

Accepted: 13-Dec-2022

Published: 16-Dec-2022

Original Article | Volume 5 - Issue 1 | Article DOI :

Download PDF

Stanislav Urazov¹*, Alina Cherkas¹,², Alexey Boykov², Sergey Shcherbak¹,³, Anna Anisenkova¹, Sergey Mosenko¹, Oleg Glotov¹,⁴, Sergey Azarenko¹, Marina Bolsunovskaya², Oleg Popov¹, and Natalya Klenkova¹

¹City Hospital 40 of Saint Petersburg, Russia

²Laboratory of Industrial Systems for Streaming Data Processing of the SPbPU National Technology Initiative Center for Advanced Manufacturing Technologies, St. Petersburg, Russia

³Federal State Budgetary Educational Institution of Higher Professional Education, Saint-Petersburg State University, St. Petersburg, Russia

?Federal State-Financed Institution Pediatric Research and Clinical Center for Infectious Diseases under the Federal Medical Biological Agency, St. Petersburg, Russia

Corresponding Author:

Stanislav Urazov, City Hospital 40 of Saint Petersburg, Russia, Tel: +790-627-664-53; ORCiD: 0000-0002-5441-2911

Keywords

COVID-19; Classifier; Risk Meter; Coronavirus Infection

Abstract

The clinical spectrum of COVID-19 ranges from asymptomatic disease to pneumonia and life-threatening complications, including acute respiratory distress syndrome, organ failure and death [1-3]. In this regard, the development of models that allow medical workers to quickly assess the likelihood of an unfavorable development of COVID-19 seems to be an extremely urgent task. The aim of this study is to develop a model for predicting the severity of the course of COVID-19. For training and validation of 19 machine learning models, 117 clinical and laboratory parameters were used for 10487 patients with coronavirus infection who were treated at City Hospital No 40 of St. Petersburg, Russia from 01.09.2020 to 15.10.2021. As a result, 2 best models were obtained, including 21 and 10 features with AUC = 0.91 ± 0.01 and 0.86 ± 0.01, respectively. This paper provides an extensive overview of the available models for predicting the severity of COVID-19 disease and proposes 2 developed models. A mobile application has been created for the convenience of accumulating new data and using the model.

Citation

Urazov S, Cherkas A, Boykov A, Shcherbak S, Anisenkova A, et al. (2022) Classifier for Prediction of the Level of Disease COVID-19. SM J Infect Dis 5: 12.