Your browser doesn't support javascript.
Data Mining Models for Automatic Problem Identification in Intensive Medicine.
Quesado, Inês; Duarte, Julio; Silva, Álvaro; Manuel, Maria; Quintas, César.
  • Quesado I; Algoritmi/LASI research center, University of Minho, Portugal.
  • Duarte J; Algoritmi/LASI research center, University of Minho, Portugal.
  • Silva Á; Centro Hospitalar Universitário do Porto, Portugal.
  • Manuel M; Centro Hospitalar Universitário do Porto, Portugal.
  • Quintas C; Centro Hospitalar Universitário do Porto, Portugal.
Procedia Comput Sci ; 210: 218-223, 2022.
Article in English | MEDLINE | ID: covidwho-2132118
ABSTRACT
This paper aims to support medical decision making on predicting the diagnosis of COVID-19. Thus, a set of Data Mining (DM) models was developed using prediction techniques and classification models. These models try to understand whether the vital signs of patients have a correlation with a diagnosis. To achieve the objective of the paper, initially, the data was acquired and collected from several data sources such as bedside monitors and electronic nursing records from the Intensive Care Unit of the Santo António Hospital. Secondly, the data was transformed so that it could be used in DM models. The models were induced using the following algorithms Decision Trees, Random Forest, Naive Bayes, and Support Vector Machine. The analysis of the sensitivity, specificity, and accuracy were the metrics used to identify the most relevant measures to predict COVID-19 diagnosis. This work demonstrates that the models created had promising results.
Keywords

Full text: Available Collection: International databases Database: MEDLINE Type of study: Diagnostic study / Prognostic study / Randomized controlled trials / Reviews Language: English Journal: Procedia Comput Sci Year: 2022 Document Type: Article Affiliation country: J.procs.2022.10.140

Similar

MEDLINE

...
LILACS

LIS


Full text: Available Collection: International databases Database: MEDLINE Type of study: Diagnostic study / Prognostic study / Randomized controlled trials / Reviews Language: English Journal: Procedia Comput Sci Year: 2022 Document Type: Article Affiliation country: J.procs.2022.10.140