Publicação

Uncertainty-Aware AI for ECG arrhythmia multi-label classification

Ver documento

Detalhes bibliográficos
Resumo:Machine Learning (ML) models are able to predict a variety of diseases, with performances that can be superior to those achieved by healthcare professionals. However, when implemented in clinical settings as decision support systems, their generalisation capabilities are often compromised, rendering healthcare professionals more susceptible into delivering erroneous diagnostics. This research focuses on uncertainty measures as a key method to abstain from classifying samples with high uncertainty as well as a selection criterion for active learning strategies. For this purpose, it was employed four large public multi-label Electrocardiogram (ECG) databases for the classification of cardiac arrhythmias. Regarding the uncertainty measures, single distribution uncertainty and classical information-theoretic measures of entropy were tested and compared. Thus, three Deep Learning models were developed: a single convolutional neural network and two multiple-models using Monte-Carlo Dropout and Deep Ensemble techniques. When tested with samples from the same database used for training, all models achieved performances higher than 95% for F1-score. However, when tested on an external dataset, their performances dropped to approximately 70%, indicating a probable scenario of dataset shift. The Deep Ensemble model obtained the highest F1-score in both test sets with a maximum difference of 3% from the others. The classification withrejection option increased from a rejection of10% to a range between 30% to 50% depending on the model or uncertainty measure, with the highest rejection rates being obtained on external data. This reveals that external dataset’s classifications have higher uncertainty, also an indication of dataset shift. For the active learning approach, 10% of the highest uncertainty sampleswere used to retrain the models. The performances results increased by almost 5%, suggesting uncertainty as a good selection method. Although there are still challenges to the implementation of ML models, the preliminary studies show that uncertainty quantification is a valuable method for classification with rejection option and active learning approaches under dataset shift conditions.
Autores principais:Simão, Raquel Filipa Birra
Assunto:Uncertainty Quantification Monte Carlo Dropout Deep Ensemble Dataset shift Active Learning
Ano:2022
País:Portugal
Tipo de documento:dissertação de mestrado
Tipo de acesso:acesso aberto
Instituição associada:Universidade Nova de Lisboa
Idioma:inglês
Origem:Repositório Institucional da UNL
_version_ 1865920612049879040
author Simão, Raquel Filipa Birra
author_facet Simão, Raquel Filipa Birra
Simão, Raquel Filipa Birra
author_role author
contributor_name_str_mv Gamboa, Hugo
RUN
country_str PT
creators_json_str [{\"Person.name\":\"Simão, Raquel Filipa Birra\"}]
datacite.contributors.contributor.contributorName.fl_str_mv Gamboa, Hugo
RUN
datacite.creators.creator.creatorName.fl_str_mv Simão, Raquel Filipa Birra
datacite.date.Accepted.fl_str_mv 2022-11-01T00:00:00Z
datacite.date.available.fl_str_mv 2023-09-01T13:03:03Z
datacite.date.embargoed.fl_str_mv 2023-09-01T13:03:03Z
datacite.rights.fl_str_mv http://purl.org/coar/access_right/c_abf2
datacite.subjects.subject.fl_str_mv Uncertainty Quantification
Monte Carlo Dropout
Deep Ensemble
Dataset shift
Active Learning
datacite.titles.title.fl_str_mv Uncertainty-Aware AI for ECG arrhythmia multi-label classification
dc.contributor.none.fl_str_mv Gamboa, Hugo
RUN
dc.creator.none.fl_str_mv Simão, Raquel Filipa Birra
dc.date.Accepted.fl_str_mv 2022-11-01T00:00:00Z
dc.date.available.fl_str_mv 2023-09-01T13:03:03Z
dc.date.embargoed.fl_str_mv 2023-09-01T13:03:03Z
dc.format.none.fl_str_mv application/pdf
dc.identifier.none.fl_str_mv http://hdl.handle.net/10362/157128
dc.language.none.fl_str_mv eng
dc.rights.none.fl_str_mv http://purl.org/coar/access_right/c_abf2
dc.subject.none.fl_str_mv Uncertainty Quantification
Monte Carlo Dropout
Deep Ensemble
Dataset shift
Active Learning
dc.title.fl_str_mv Uncertainty-Aware AI for ECG arrhythmia multi-label classification
dc.type.none.fl_str_mv http://purl.org/coar/resource_type/c_bdcc
description Machine Learning (ML) models are able to predict a variety of diseases, with performances that can be superior to those achieved by healthcare professionals. However, when implemented in clinical settings as decision support systems, their generalisation capabilities are often compromised, rendering healthcare professionals more susceptible into delivering erroneous diagnostics. This research focuses on uncertainty measures as a key method to abstain from classifying samples with high uncertainty as well as a selection criterion for active learning strategies. For this purpose, it was employed four large public multi-label Electrocardiogram (ECG) databases for the classification of cardiac arrhythmias. Regarding the uncertainty measures, single distribution uncertainty and classical information-theoretic measures of entropy were tested and compared. Thus, three Deep Learning models were developed: a single convolutional neural network and two multiple-models using Monte-Carlo Dropout and Deep Ensemble techniques. When tested with samples from the same database used for training, all models achieved performances higher than 95% for F1-score. However, when tested on an external dataset, their performances dropped to approximately 70%, indicating a probable scenario of dataset shift. The Deep Ensemble model obtained the highest F1-score in both test sets with a maximum difference of 3% from the others. The classification withrejection option increased from a rejection of10% to a range between 30% to 50% depending on the model or uncertainty measure, with the highest rejection rates being obtained on external data. This reveals that external dataset’s classifications have higher uncertainty, also an indication of dataset shift. For the active learning approach, 10% of the highest uncertainty sampleswere used to retrain the models. The performances results increased by almost 5%, suggesting uncertainty as a good selection method. Although there are still challenges to the implementation of ML models, the preliminary studies show that uncertainty quantification is a valuable method for classification with rejection option and active learning approaches under dataset shift conditions.
dirty 0
eu_rights_str_mv openAccess
format masterThesis
fulltext.url.fl_str_mv https://run.unl.pt/bitstreams/25ee8c28-95e7-4a26-8b7e-aa2521cefeaa/download
id run_4e3bd52bcffecb83eb8a6960698022ba
identifier.url.fl_str_mv http://hdl.handle.net/10362/157128
instacron_str unl
institution Universidade Nova de Lisboa
instname_str Universidade Nova de Lisboa
language eng
network_acronym_str run
network_name_str Repositório Institucional da UNL
oai_identifier_str oai:run.unl.pt:10362/157128
organization_str_mv urn:organizationAcronym:unl
person_str_mv Simão, Raquel Filipa Birra
publishDate 2022
reponame_str Repositório Institucional da UNL
repository_id_str urn:repositoryAcronym:run
service_str_mv urn:repositoryAcronym:run
spelling engpt_PTMachine Learning (ML) models are able to predict a variety of diseases, with performances that can be superior to those achieved by healthcare professionals. However, when implemented in clinical settings as decision support systems, their generalisation capabilities are often compromised, rendering healthcare professionals more susceptible into delivering erroneous diagnostics. This research focuses on uncertainty measures as a key method to abstain from classifying samples with high uncertainty as well as a selection criterion for active learning strategies. For this purpose, it was employed four large public multi-label Electrocardiogram (ECG) databases for the classification of cardiac arrhythmias. Regarding the uncertainty measures, single distribution uncertainty and classical information-theoretic measures of entropy were tested and compared. Thus, three Deep Learning models were developed: a single convolutional neural network and two multiple-models using Monte-Carlo Dropout and Deep Ensemble techniques. When tested with samples from the same database used for training, all models achieved performances higher than 95% for F1-score. However, when tested on an external dataset, their performances dropped to approximately 70%, indicating a probable scenario of dataset shift. The Deep Ensemble model obtained the highest F1-score in both test sets with a maximum difference of 3% from the others. The classification withrejection option increased from a rejection of10% to a range between 30% to 50% depending on the model or uncertainty measure, with the highest rejection rates being obtained on external data. This reveals that external dataset’s classifications have higher uncertainty, also an indication of dataset shift. For the active learning approach, 10% of the highest uncertainty sampleswere used to retrain the models. The performances results increased by almost 5%, suggesting uncertainty as a good selection method. Although there are still challenges to the implementation of ML models, the preliminary studies show that uncertainty quantification is a valuable method for classification with rejection option and active learning approaches under dataset shift conditions.application/pdfpt_PTUncertainty-Aware AI for ECG arrhythmia multi-label classificationSimão, Raquel Filipa BirraGamboa, HugoHostingInstitutionOrganizationalRUNe-mailmailto:run@unl.ptrun@unl.pt2023-09-01T13:03:03Z2022-112022-11-01T00:00:00ZHandlehttp://hdl.handle.net/10362/157128http://purl.org/coar/access_right/c_abf2open accessUncertainty QuantificationMonte Carlo DropoutDeep EnsembleDataset shiftActive Learning15763171 bytesliteraturehttp://purl.org/coar/resource_type/c_bdccmaster thesishttp://purl.org/coar/access_right/c_abf2application/pdffulltexthttps://run.unl.pt/bitstreams/25ee8c28-95e7-4a26-8b7e-aa2521cefeaa/download
spellingShingle Uncertainty-Aware AI for ECG arrhythmia multi-label classification
Uncertainty-Aware AI for ECG arrhythmia multi-label classification
Simão, Raquel Filipa Birra
Uncertainty Quantification
Monte Carlo Dropout
Deep Ensemble
Dataset shift
Active Learning
Simão, Raquel Filipa Birra
Uncertainty Quantification
Monte Carlo Dropout
Deep Ensemble
Dataset shift
Active Learning
status NEW
subject.fl_str_mv Uncertainty Quantification
Monte Carlo Dropout
Deep Ensemble
Dataset shift
Active Learning
title Uncertainty-Aware AI for ECG arrhythmia multi-label classification
title_full Uncertainty-Aware AI for ECG arrhythmia multi-label classification
title_fullStr Uncertainty-Aware AI for ECG arrhythmia multi-label classification
Uncertainty-Aware AI for ECG arrhythmia multi-label classification
title_full_unstemmed Uncertainty-Aware AI for ECG arrhythmia multi-label classification
Uncertainty-Aware AI for ECG arrhythmia multi-label classification
title_short Uncertainty-Aware AI for ECG arrhythmia multi-label classification
title_sort Uncertainty-Aware AI for ECG arrhythmia multi-label classification
topic Uncertainty Quantification
Monte Carlo Dropout
Deep Ensemble
Dataset shift
Active Learning
topic_facet Uncertainty Quantification
Monte Carlo Dropout
Deep Ensemble
Dataset shift
Active Learning
url http://hdl.handle.net/10362/157128
visible 1