Prediction of Early TBI Mortality Using a Machine Learning Approach in a LMIC Population

Vista sencilla de ítem
dc.contributor.authorAmorim, Robson
dc.contributor.authorOliveira, Louise
dc.contributor.authorMalbouisson, Luiz Marcelo
dc.contributor.authorNagumo, Marcia
dc.contributor.authorSimoes, Marcela
dc.contributor.authorBor-Seng-Shu, Edson
dc.contributor.authorBeer-Furlan, André
dc.contributor.authorFerreira de Andrade, Almir
dc.contributor.authorRubiano, Andrés M.
dc.contributor.authorTeixeira, Manoel Jacobsen
dc.contributor.authorKolias, Angelos
dc.contributor.authorPaiva, Vera
dc.contributor.orcidRubiano, Andrés M. [0000-0001-8931-3254]
dc.date.accessioned2020-03-07T15:42:42Z
dc.date.available2020-03-07T15:42:42Z
dc.date.issued2020
dc.description.abstractenglishBackground: In a time when the incidence of severe traumatic brain injury (TBI) is increasing in low- to middle-income countries (LMICs), it is important to understand the behavior of predictive variables in an LMIC’s population. There are few previous attempts to generate prediction models for TBI outcomes from local data in LMICs. Our study aim is to design and compare a series of predictive models for mortality on a new cohort in TBI patients in Brazil using Machine Learning. Methods: A prospective registry was set in São Paulo, Brazil, enrolling all patients with a diagnosis of TBI that require admission to the intensive care unit. We evaluated the following predictors: gender, age, pupil reactivity at admission, Glasgow Coma Scale (GCS), presence of hypoxia and hypotension, computed tomography findings, trauma severity score, and laboratory results. Results: Overall mortality at 14 days was 22.8%. Models had a high prediction performance, with the best prediction for overall mortality achieved through Naive Bayes (area under the curve = 0.906). The most significant predictors were the GCS at admission and prehospital GCS, age, and pupil reaction. When predicting the length of stay at the intensive care unit, the Conditional Inference Tree model had the best performance (root mean square error = 1.011), with the most important variable across all models being the GCS at scene. Conclusions: Models for early mortality and hospital length of stay using Machine Learning can achieve high performance when based on registry data even in LMICs. These models have the potential to inform treatment decisions and counsel family members. Level of evidence: This observational study provides a level IV evidence on prognosis after TBI.eng
dc.format.mimetypeapplication/pdf
dc.identifier.doihttps://doi.org/10.3389/fneur.2019.01366
dc.identifier.instnameinstname:Universidad El Bosquespa
dc.identifier.issn1664-2295
dc.identifier.reponamereponame:Repositorio Institucional Universidad El Bosquespa
dc.identifier.repourlrepourl:https://repositorio.unbosque.edu.co
dc.identifier.urihttps://hdl.handle.net/20.500.12495/2014
dc.language.isoeng
dc.publisherFrontiers Media S.A.spa
dc.publisher.journalFrontiers in neurologyspa
dc.relation.ispartofseriesFrontiers in neurology, 1664-2295, Vol. 10, 2020spa
dc.relation.urihttps://www.frontiersin.org/articles/10.3389/fneur.2019.01366/full
dc.rightsAttribution 4.0 International*
dc.rights.accessrightsinfo:eu-repo/semantics/openAccess
dc.rights.accessrightshttps://purl.org/coar/access_right/c_abf437
dc.rights.creativecommons2020
dc.rights.localAcceso abiertospa
dc.rights.urihttps://creativecommons.org/licenses/by/4.0/*
dc.subject.decsTomografía computarizada espiralspa
dc.subject.decsEscala de coma de glasgowspa
dc.subject.decsPruebas diagnósticas de rutinaspa
dc.subject.keywordsLMICsspa
dc.subject.keywordsMachine learningspa
dc.subject.keywordsMortalityspa
dc.titlePrediction of Early TBI Mortality Using a Machine Learning Approach in a LMIC Populationspa
dc.typearticlespa
dc.type.hasversioninfo:eu-repo/semantics/publishedVersion
dc.type.localartículospa

Archivos

Bloque original
Mostrando 1 - 1 de 1
Miniatura
Nombre:
Amorim, R.L._2020.pdf
Tamaño:
1.15 MB
Formato:
Adobe Portable Document Format
Descripción:
Descargar
Bloque de licencias
Mostrando 1 - 1 de 1
No hay miniatura disponible
Nombre:
license.txt
Tamaño:
1.71 KB
Formato:
Item-specific license agreed upon to submission
Descripción:
Descargar

Colecciones

Artículos