Magnetic Resonance Sequences: Experimental Assessment of Achievements and Limitations
Autor(a) principal: | |
---|---|
Data de Publicação: | 2021 |
Tipo de documento: | Artigo |
Idioma: | eng |
Título da fonte: | Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
Texto Completo: | http://hdl.handle.net/10316/100892 https://doi.org/10.12720/jait.12.1.66-70 |
Resumo: | Deep Learning can be applied to learn segmentations of abdominal organs in MRI sequences, a challenging task due to changing morphologies of organs along different slices. Evaluation of outcome is important to decide on applicability and to command further improvements. Software tools include evaluation metrics. Some metrics indicate quasi-perfection, with potential erroneous conclusions, visual inspection and some per organ metrics say otherwise. Our aim is the correct interpretation of commonly available metrics on organs segmentation. The method to do that is to build two architectures (DeepLab, FCN), run segmentation experiments, interpret results. Examples of results as aggregates (mean accuracy 98% weighted IoU 97%) are overly optimistic. Further analysis shows much lower scores (mean IoU 68% IoU of individual organs 78, 66, 59, 41%). We conclude that correct interpretation of the metrics, importance of further architectural or post-processing improvements on false positives. |
id |
RCAP_2e21b7c804e510b56ca9854c7011c415 |
---|---|
oai_identifier_str |
oai:estudogeral.uc.pt:10316/100892 |
network_acronym_str |
RCAP |
network_name_str |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
repository_id_str |
7160 |
spelling |
Magnetic Resonance Sequences: Experimental Assessment of Achievements and Limitationssegmentationdeep learningassessmentDeep Learning can be applied to learn segmentations of abdominal organs in MRI sequences, a challenging task due to changing morphologies of organs along different slices. Evaluation of outcome is important to decide on applicability and to command further improvements. Software tools include evaluation metrics. Some metrics indicate quasi-perfection, with potential erroneous conclusions, visual inspection and some per organ metrics say otherwise. Our aim is the correct interpretation of commonly available metrics on organs segmentation. The method to do that is to build two architectures (DeepLab, FCN), run segmentation experiments, interpret results. Examples of results as aggregates (mean accuracy 98% weighted IoU 97%) are overly optimistic. Further analysis shows much lower scores (mean IoU 68% IoU of individual organs 78, 66, 59, 41%). We conclude that correct interpretation of the metrics, importance of further architectural or post-processing improvements on false positives.2021info:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/articlehttp://hdl.handle.net/10316/100892http://hdl.handle.net/10316/100892https://doi.org/10.12720/jait.12.1.66-70eng17982340Furtado, Pedroinfo: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:RCAAP2022-07-19T20:37:40Zoai:estudogeral.uc.pt:10316/100892Portal AgregadorONGhttps://www.rcaap.pt/oai/openaireopendoar:71602024-03-19T21:18:11.069174Repositório 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 |
Magnetic Resonance Sequences: Experimental Assessment of Achievements and Limitations |
title |
Magnetic Resonance Sequences: Experimental Assessment of Achievements and Limitations |
spellingShingle |
Magnetic Resonance Sequences: Experimental Assessment of Achievements and Limitations Furtado, Pedro segmentation deep learning assessment |
title_short |
Magnetic Resonance Sequences: Experimental Assessment of Achievements and Limitations |
title_full |
Magnetic Resonance Sequences: Experimental Assessment of Achievements and Limitations |
title_fullStr |
Magnetic Resonance Sequences: Experimental Assessment of Achievements and Limitations |
title_full_unstemmed |
Magnetic Resonance Sequences: Experimental Assessment of Achievements and Limitations |
title_sort |
Magnetic Resonance Sequences: Experimental Assessment of Achievements and Limitations |
author |
Furtado, Pedro |
author_facet |
Furtado, Pedro |
author_role |
author |
dc.contributor.author.fl_str_mv |
Furtado, Pedro |
dc.subject.por.fl_str_mv |
segmentation deep learning assessment |
topic |
segmentation deep learning assessment |
description |
Deep Learning can be applied to learn segmentations of abdominal organs in MRI sequences, a challenging task due to changing morphologies of organs along different slices. Evaluation of outcome is important to decide on applicability and to command further improvements. Software tools include evaluation metrics. Some metrics indicate quasi-perfection, with potential erroneous conclusions, visual inspection and some per organ metrics say otherwise. Our aim is the correct interpretation of commonly available metrics on organs segmentation. The method to do that is to build two architectures (DeepLab, FCN), run segmentation experiments, interpret results. Examples of results as aggregates (mean accuracy 98% weighted IoU 97%) are overly optimistic. Further analysis shows much lower scores (mean IoU 68% IoU of individual organs 78, 66, 59, 41%). We conclude that correct interpretation of the metrics, importance of further architectural or post-processing improvements on false positives. |
publishDate |
2021 |
dc.date.none.fl_str_mv |
2021 |
dc.type.status.fl_str_mv |
info:eu-repo/semantics/publishedVersion |
dc.type.driver.fl_str_mv |
info:eu-repo/semantics/article |
format |
article |
status_str |
publishedVersion |
dc.identifier.uri.fl_str_mv |
http://hdl.handle.net/10316/100892 http://hdl.handle.net/10316/100892 https://doi.org/10.12720/jait.12.1.66-70 |
url |
http://hdl.handle.net/10316/100892 https://doi.org/10.12720/jait.12.1.66-70 |
dc.language.iso.fl_str_mv |
eng |
language |
eng |
dc.relation.none.fl_str_mv |
17982340 |
dc.rights.driver.fl_str_mv |
info:eu-repo/semantics/openAccess |
eu_rights_str_mv |
openAccess |
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_ |
1799134076897067008 |