Linking patient data to scientific knowledge to support contextualized mining
Autor(a) principal: | |
---|---|
Data de Publicação: | 2022 |
Tipo de documento: | Dissertação |
Idioma: | eng |
Título da fonte: | Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
Texto Completo: | http://hdl.handle.net/10451/53897 |
Resumo: | Tese de mestrado, Bioinformática e Biologia Computacional, Universidade de Lisboa, Faculdade de Ciências, 2022 |
id |
RCAP_596100195e2fdbe2453d8135f9b2a1b8 |
---|---|
oai_identifier_str |
oai:repositorio.ul.pt:10451/53897 |
network_acronym_str |
RCAP |
network_name_str |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
repository_id_str |
7160 |
spelling |
Linking patient data to scientific knowledge to support contextualized miningAnotação semanticaOntologias BiomédicasReadmissões em UCIApredizagme automáticaEmbedding de grafos de conhecimentoTeses de mestrado - 2022Departamento de InformáticaTese de mestrado, Bioinformática e Biologia Computacional, Universidade de Lisboa, Faculdade de Ciências, 2022ICU readmissions are a critical problem associated with either serious conditions, ill nesses, or complications, representing a 4 times increase in mortality risk and a financial burden to health institutions. In developed countries 1 in every 10 patients discharged comes back to the ICU. As hospitals become more and more data-oriented with the adop tion of Electronic Health Records (EHR), there as been a rise in the development of com putational approaches to support clinical decision. In recent years new efforts emerged, using machine learning approaches to make ICU readmission predictions directly over EHR data. Despite these growing efforts, machine learning approaches still explore EHR data directly without taking into account its mean ing or context. Medical knowledge is not accessible to these methods, who work blindly over the data, without considering the meaning and relationships the data objects. Ontolo gies and knowledge graphs can help bridge this gap between data and scientific context, since they are computational artefacts that represent the entities in a domain and how the relate to each other in a formalized fashion. This opportunity motivated the aim of this work: to investigate how enriching EHR data with ontology-based semantic annotations and applying machine learning techniques that explore them can impact the prediction of 30-day ICU readmission risk. To achieve this, a number of contributions were developed, including: (1) An enrichment of the MIMIC-III data set with annotations to several biomedical ontologies; (2) A novel ap proach to predict ICU readmission risk that explores knowledge graph embeddings to represent patient data taking into account the semantic annotations; (3) A variant of the predictive approach that targets different moments to support risk prediction throughout the ICU stay. The predictive approaches outperformed both state-of-the-art and a baseline achieving a ROC-AUC of 0.815 (an increase of 0.2 over the state of the art). The positive results achieved motivated the development of an entrepreneurial project, which placed in the Top 5 of the H-INNOVA 2021 entrepreneurship award.Pesquita, Cátia, 1980-Oliveira, Daniela Patrícia dos SantosRepositório da Universidade de LisboaCarvalho, Ricardo Miguel Serafim de2022-07-21T09:05:38Z202220222022-01-01T00:00:00Zinfo:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/masterThesisapplication/pdfhttp://hdl.handle.net/10451/53897enginfo:eu-repo/semantics/openAccessreponame:Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)instname:Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informaçãoinstacron:RCAAP2023-11-08T17:00:08Zoai:repositorio.ul.pt:10451/53897Portal AgregadorONGhttps://www.rcaap.pt/oai/openaireopendoar:71602024-03-19T22:04:53.036845Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) - Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informaçãofalse |
dc.title.none.fl_str_mv |
Linking patient data to scientific knowledge to support contextualized mining |
title |
Linking patient data to scientific knowledge to support contextualized mining |
spellingShingle |
Linking patient data to scientific knowledge to support contextualized mining Carvalho, Ricardo Miguel Serafim de Anotação semantica Ontologias Biomédicas Readmissões em UCI Apredizagme automática Embedding de grafos de conhecimento Teses de mestrado - 2022 Departamento de Informática |
title_short |
Linking patient data to scientific knowledge to support contextualized mining |
title_full |
Linking patient data to scientific knowledge to support contextualized mining |
title_fullStr |
Linking patient data to scientific knowledge to support contextualized mining |
title_full_unstemmed |
Linking patient data to scientific knowledge to support contextualized mining |
title_sort |
Linking patient data to scientific knowledge to support contextualized mining |
author |
Carvalho, Ricardo Miguel Serafim de |
author_facet |
Carvalho, Ricardo Miguel Serafim de |
author_role |
author |
dc.contributor.none.fl_str_mv |
Pesquita, Cátia, 1980- Oliveira, Daniela Patrícia dos Santos Repositório da Universidade de Lisboa |
dc.contributor.author.fl_str_mv |
Carvalho, Ricardo Miguel Serafim de |
dc.subject.por.fl_str_mv |
Anotação semantica Ontologias Biomédicas Readmissões em UCI Apredizagme automática Embedding de grafos de conhecimento Teses de mestrado - 2022 Departamento de Informática |
topic |
Anotação semantica Ontologias Biomédicas Readmissões em UCI Apredizagme automática Embedding de grafos de conhecimento Teses de mestrado - 2022 Departamento de Informática |
description |
Tese de mestrado, Bioinformática e Biologia Computacional, Universidade de Lisboa, Faculdade de Ciências, 2022 |
publishDate |
2022 |
dc.date.none.fl_str_mv |
2022-07-21T09:05:38Z 2022 2022 2022-01-01T00:00:00Z |
dc.type.status.fl_str_mv |
info:eu-repo/semantics/publishedVersion |
dc.type.driver.fl_str_mv |
info:eu-repo/semantics/masterThesis |
format |
masterThesis |
status_str |
publishedVersion |
dc.identifier.uri.fl_str_mv |
http://hdl.handle.net/10451/53897 |
url |
http://hdl.handle.net/10451/53897 |
dc.language.iso.fl_str_mv |
eng |
language |
eng |
dc.rights.driver.fl_str_mv |
info:eu-repo/semantics/openAccess |
eu_rights_str_mv |
openAccess |
dc.format.none.fl_str_mv |
application/pdf |
dc.source.none.fl_str_mv |
reponame:Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) instname:Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informação instacron:RCAAP |
instname_str |
Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informação |
instacron_str |
RCAAP |
institution |
RCAAP |
reponame_str |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
collection |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
repository.name.fl_str_mv |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) - Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informação |
repository.mail.fl_str_mv |
|
_version_ |
1799134600382906368 |