Cargando…
A machine learning based exploration of COVID-19 mortality risk
Early prediction of patient mortality risks during a pandemic can decrease mortality by assuring efficient resource allocation and treatment planning. This study aimed to develop and compare prognosis prediction machine learning models based on invasive laboratory and noninvasive clinical and demogr...
Autores principales: | , , , , , , , , , , , |
---|---|
Formato: | Online Artículo Texto |
Lenguaje: | English |
Publicado: |
Public Library of Science
2021
|
Materias: | |
Acceso en línea: | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8253432/ https://www.ncbi.nlm.nih.gov/pubmed/34214101 http://dx.doi.org/10.1371/journal.pone.0252384 |
_version_ | 1783717511173242880 |
---|---|
author | Mahdavi, Mahdi Choubdar, Hadi Zabeh, Erfan Rieder, Michael Safavi-Naeini, Safieddin Jobbagy, Zsolt Ghorbani, Amirata Abedini, Atefeh Kiani, Arda Khanlarzadeh, Vida Lashgari, Reza Kamrani, Ehsan |
author_facet | Mahdavi, Mahdi Choubdar, Hadi Zabeh, Erfan Rieder, Michael Safavi-Naeini, Safieddin Jobbagy, Zsolt Ghorbani, Amirata Abedini, Atefeh Kiani, Arda Khanlarzadeh, Vida Lashgari, Reza Kamrani, Ehsan |
author_sort | Mahdavi, Mahdi |
collection | PubMed |
description | Early prediction of patient mortality risks during a pandemic can decrease mortality by assuring efficient resource allocation and treatment planning. This study aimed to develop and compare prognosis prediction machine learning models based on invasive laboratory and noninvasive clinical and demographic data from patients’ day of admission. Three Support Vector Machine (SVM) models were developed and compared using invasive, non-invasive, and both groups. The results suggested that non-invasive features could provide mortality predictions that are similar to the invasive and roughly on par with the joint model. Feature inspection results from SVM-RFE and sparsity analysis displayed that, compared with the invasive model, the non-invasive model can provide better performances with a fewer number of features, pointing to the presence of high predictive information contents in several non-invasive features, including SPO(2), age, and cardiovascular disorders. Furthermore, while the invasive model was able to provide better mortality predictions for the imminent future, non-invasive features displayed better performance for more distant expiration intervals. Early mortality prediction using non-invasive models can give us insights as to where and with whom to intervene. Combined with novel technologies, such as wireless wearable devices, these models can create powerful frameworks for various medical assignments and patient triage. |
format | Online Article Text |
id | pubmed-8253432 |
institution | National Center for Biotechnology Information |
language | English |
publishDate | 2021 |
publisher | Public Library of Science |
record_format | MEDLINE/PubMed |
spelling | pubmed-82534322021-07-13 A machine learning based exploration of COVID-19 mortality risk Mahdavi, Mahdi Choubdar, Hadi Zabeh, Erfan Rieder, Michael Safavi-Naeini, Safieddin Jobbagy, Zsolt Ghorbani, Amirata Abedini, Atefeh Kiani, Arda Khanlarzadeh, Vida Lashgari, Reza Kamrani, Ehsan PLoS One Research Article Early prediction of patient mortality risks during a pandemic can decrease mortality by assuring efficient resource allocation and treatment planning. This study aimed to develop and compare prognosis prediction machine learning models based on invasive laboratory and noninvasive clinical and demographic data from patients’ day of admission. Three Support Vector Machine (SVM) models were developed and compared using invasive, non-invasive, and both groups. The results suggested that non-invasive features could provide mortality predictions that are similar to the invasive and roughly on par with the joint model. Feature inspection results from SVM-RFE and sparsity analysis displayed that, compared with the invasive model, the non-invasive model can provide better performances with a fewer number of features, pointing to the presence of high predictive information contents in several non-invasive features, including SPO(2), age, and cardiovascular disorders. Furthermore, while the invasive model was able to provide better mortality predictions for the imminent future, non-invasive features displayed better performance for more distant expiration intervals. Early mortality prediction using non-invasive models can give us insights as to where and with whom to intervene. Combined with novel technologies, such as wireless wearable devices, these models can create powerful frameworks for various medical assignments and patient triage. Public Library of Science 2021-07-02 /pmc/articles/PMC8253432/ /pubmed/34214101 http://dx.doi.org/10.1371/journal.pone.0252384 Text en © 2021 Mahdavi et al https://creativecommons.org/licenses/by/4.0/This is an open access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/) , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. |
spellingShingle | Research Article Mahdavi, Mahdi Choubdar, Hadi Zabeh, Erfan Rieder, Michael Safavi-Naeini, Safieddin Jobbagy, Zsolt Ghorbani, Amirata Abedini, Atefeh Kiani, Arda Khanlarzadeh, Vida Lashgari, Reza Kamrani, Ehsan A machine learning based exploration of COVID-19 mortality risk |
title | A machine learning based exploration of COVID-19 mortality risk |
title_full | A machine learning based exploration of COVID-19 mortality risk |
title_fullStr | A machine learning based exploration of COVID-19 mortality risk |
title_full_unstemmed | A machine learning based exploration of COVID-19 mortality risk |
title_short | A machine learning based exploration of COVID-19 mortality risk |
title_sort | machine learning based exploration of covid-19 mortality risk |
topic | Research Article |
url | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8253432/ https://www.ncbi.nlm.nih.gov/pubmed/34214101 http://dx.doi.org/10.1371/journal.pone.0252384 |
work_keys_str_mv | AT mahdavimahdi amachinelearningbasedexplorationofcovid19mortalityrisk AT choubdarhadi amachinelearningbasedexplorationofcovid19mortalityrisk AT zabeherfan amachinelearningbasedexplorationofcovid19mortalityrisk AT riedermichael amachinelearningbasedexplorationofcovid19mortalityrisk AT safavinaeinisafieddin amachinelearningbasedexplorationofcovid19mortalityrisk AT jobbagyzsolt amachinelearningbasedexplorationofcovid19mortalityrisk AT ghorbaniamirata amachinelearningbasedexplorationofcovid19mortalityrisk AT abediniatefeh amachinelearningbasedexplorationofcovid19mortalityrisk AT kianiarda amachinelearningbasedexplorationofcovid19mortalityrisk AT khanlarzadehvida amachinelearningbasedexplorationofcovid19mortalityrisk AT lashgarireza amachinelearningbasedexplorationofcovid19mortalityrisk AT kamraniehsan amachinelearningbasedexplorationofcovid19mortalityrisk AT mahdavimahdi machinelearningbasedexplorationofcovid19mortalityrisk AT choubdarhadi machinelearningbasedexplorationofcovid19mortalityrisk AT zabeherfan machinelearningbasedexplorationofcovid19mortalityrisk AT riedermichael machinelearningbasedexplorationofcovid19mortalityrisk AT safavinaeinisafieddin machinelearningbasedexplorationofcovid19mortalityrisk AT jobbagyzsolt machinelearningbasedexplorationofcovid19mortalityrisk AT ghorbaniamirata machinelearningbasedexplorationofcovid19mortalityrisk AT abediniatefeh machinelearningbasedexplorationofcovid19mortalityrisk AT kianiarda machinelearningbasedexplorationofcovid19mortalityrisk AT khanlarzadehvida machinelearningbasedexplorationofcovid19mortalityrisk AT lashgarireza machinelearningbasedexplorationofcovid19mortalityrisk AT kamraniehsan machinelearningbasedexplorationofcovid19mortalityrisk |