Download PDFOpen PDF in browser

Predicting Intensive Care Unit (ICU) Requirement in COVID-19 With Artificial Neural Network

EasyChair Preprint no. 5753

10 pagesDate: June 7, 2021

Abstract

Due to the fact that the first pneumonia case caused by the new 2019 coronavirus (COVID -19) was found in Wuhan, the impact and consequences of this epidemic, which has affected the whole world, are still ongoing. The health sector is one of the most affected sectors by this epidemic. Since the disease was new at the beginning and the treatment was naturally found later, the course of the disease has been fluctuating and is still not over. One of the greatest difficulties experienced by the health sector was the increase in hospital admissions due to the spread of the disease. How can we better manage the difficulties experienced in hospital capacities in such situations and how can we plan more efficiently has been the main question that this thesis explores and tries to find.

To this end, studies on patients with COVID-19 disease were reviewed in the literature. The common characteristics of COVID-19 patients were clarified. Demographics, symptoms of coronavirus disease 2019, laboratory evaluations, and clinical management were summarized. No geographical or other limitations were found in this study. The risk factors included in the literature are divided into four main sections. These sections are demographic, comorbidities, geographic, and lifestyle factors.

In this study, it was predicted with 80% accuracy whether a person would be admitted to the Intensive Care Unit, according to the characteristics of the artificial neural network method. Prescriptive analytics and methods are included. However, if prescriptive analytics are used to make predictions in addition to statistical analysis, a tool can be developed to assist healthcare systems and physicians.

Keyphrases: Artificial Neural Network, Big Data Analytics, COVID-19, machine learning, Prescriptive Analytics, risk factors

BibTeX entry
BibTeX does not have the right entry for preprints. This is a hack for producing the correct reference:
@Booklet{EasyChair:5753,
  author = {Yeliz Çotoy and Elif Doğu and Y. Esra Albayrak},
  title = {Predicting Intensive Care Unit (ICU) Requirement in COVID-19 With Artificial Neural Network},
  howpublished = {EasyChair Preprint no. 5753},

  year = {EasyChair, 2021}}
Download PDFOpen PDF in browser