Automatic left ventricle volume calculation with explainability through a deep learning weak-supervision methodology.

BACKGROUND AND OBJECTIVE: Magnetic resonance imaging is the most reliable imaging technique to assess the heart. More specifically there is great importance in the analysis of the left ventricle, as the main pathologies directly affect this region. In order to characterize the left ventricle, it is...

Descripción completa

Detalles Bibliográficos
Autores: Perez-Pelegri, M, Monmeneu, JV, Lopez-Lereu, MP, Perez-Pelegri, L, Maceira, AM, Bodi, V, Moratal, D
Tipo de recurso: artículo
Estado:Versión publicada
Fecha de publicación:2021
País:España
Institución:INCLIVA
Repositorio:r-INCLIVA. Repositorio Institucional de Producción Científica de INCLIVA
OAI Identifier:oai:incliva.fundanetsuite.com:p15844
Acceso en línea:https://incliva.portalinvestigacion.com/publicaciones/15844
Access Level:acceso abierto
Palabra clave:Deep learning
Explainability
Left ventricle
Magnetic resonance imaging
Segmentation
Weak supervision
id ES_7efa27f3e4246fa6b1f06f0ba99ca873
oai_identifier_str oai:incliva.fundanetsuite.com:p15844
network_acronym_str ES
network_name_str España
repository_id_str
spelling Automatic left ventricle volume calculation with explainability through a deep learning weak-supervision methodology.Perez-Pelegri, MMonmeneu, JVLopez-Lereu, MPPerez-Pelegri, LMaceira, AMBodi, VMoratal, DDeep learningExplainabilityLeft ventricleMagnetic resonance imagingSegmentationWeak supervisionBACKGROUND AND OBJECTIVE: Magnetic resonance imaging is the most reliable imaging technique to assess the heart. More specifically there is great importance in the analysis of the left ventricle, as the main pathologies directly affect this region. In order to characterize the left ventricle, it is necessary to extract its volume. In this work we present a neural network architecture that is capable of directly estimating the left ventricle volume in short axis cine Magnetic Resonance Imaging in the end-diastolic frame and provide a segmentation of the region which is the basis of the volume calculation, thus offering explainability to the estimated value. METHODS: The network was designed to directly target the volumes to estimate, not requiring any labeled segmentation on the images. The network was based on a 3D U-net with extra layers defined in a scanning module that learned features like the circularity of the objects and the volumes to estimate in a weakly-supervised manner. The only targets defined were the left ventricle volumes and the circularity of the object detected through the estimation of the p value derived from its shape. We had access to 397 cases corresponding to 397 different subjects. We randomly selected 98 cases to use as test set. RESULTS: The results show a good match between the real and estimated volumes in the test set, with a mean relative error of 8% and a mean absolute error of 9.12 ml with a Pearson correlation coefficient of 0.95. The derived segmentations obtained by the network achieved Dice coefficients with a mean value of 0.79. CONCLUSIONS: The proposed method is capable of obtaining the left ventricle volume biomarker in the end-diastole and offer an explanation of how it obtains the result in the form of a segmentation mask without the need of segmentation labels to train the algorithm, making it a potentially more trustworthy method for clinicians and a way to train neural networks more easily when segmentation labels are not readily available.ELSEVIER IRELAND LTD2021info:eu-repo/semantics/articleinfo:eu-repo/semantics/publishedVersionhttps://incliva.portalinvestigacion.com/publicaciones/15844Computer Methods and Programs in BiomedicineISSN: 01692607ISSNe: 18727565reponame:r-INCLIVA. Repositorio Institucional de Producción Científica de INCLIVAinstname:INCLIVAInglésinfo:eu-repo/semantics/openAccessoai:incliva.fundanetsuite.com:p158442026-06-07T16:35:31Z
dc.title.none.fl_str_mv Automatic left ventricle volume calculation with explainability through a deep learning weak-supervision methodology.
title Automatic left ventricle volume calculation with explainability through a deep learning weak-supervision methodology.
spellingShingle Automatic left ventricle volume calculation with explainability through a deep learning weak-supervision methodology.
Perez-Pelegri, M
Deep learning
Explainability
Left ventricle
Magnetic resonance imaging
Segmentation
Weak supervision
title_short Automatic left ventricle volume calculation with explainability through a deep learning weak-supervision methodology.
title_full Automatic left ventricle volume calculation with explainability through a deep learning weak-supervision methodology.
title_fullStr Automatic left ventricle volume calculation with explainability through a deep learning weak-supervision methodology.
title_full_unstemmed Automatic left ventricle volume calculation with explainability through a deep learning weak-supervision methodology.
title_sort Automatic left ventricle volume calculation with explainability through a deep learning weak-supervision methodology.
dc.creator.none.fl_str_mv Perez-Pelegri, M
Monmeneu, JV
Lopez-Lereu, MP
Perez-Pelegri, L
Maceira, AM
Bodi, V
Moratal, D
author Perez-Pelegri, M
author_facet Perez-Pelegri, M
Monmeneu, JV
Lopez-Lereu, MP
Perez-Pelegri, L
Maceira, AM
Bodi, V
Moratal, D
author_role author
author2 Monmeneu, JV
Lopez-Lereu, MP
Perez-Pelegri, L
Maceira, AM
Bodi, V
Moratal, D
author2_role author
author
author
author
author
author
dc.subject.none.fl_str_mv Deep learning
Explainability
Left ventricle
Magnetic resonance imaging
Segmentation
Weak supervision
topic Deep learning
Explainability
Left ventricle
Magnetic resonance imaging
Segmentation
Weak supervision
description BACKGROUND AND OBJECTIVE: Magnetic resonance imaging is the most reliable imaging technique to assess the heart. More specifically there is great importance in the analysis of the left ventricle, as the main pathologies directly affect this region. In order to characterize the left ventricle, it is necessary to extract its volume. In this work we present a neural network architecture that is capable of directly estimating the left ventricle volume in short axis cine Magnetic Resonance Imaging in the end-diastolic frame and provide a segmentation of the region which is the basis of the volume calculation, thus offering explainability to the estimated value. METHODS: The network was designed to directly target the volumes to estimate, not requiring any labeled segmentation on the images. The network was based on a 3D U-net with extra layers defined in a scanning module that learned features like the circularity of the objects and the volumes to estimate in a weakly-supervised manner. The only targets defined were the left ventricle volumes and the circularity of the object detected through the estimation of the p value derived from its shape. We had access to 397 cases corresponding to 397 different subjects. We randomly selected 98 cases to use as test set. RESULTS: The results show a good match between the real and estimated volumes in the test set, with a mean relative error of 8% and a mean absolute error of 9.12 ml with a Pearson correlation coefficient of 0.95. The derived segmentations obtained by the network achieved Dice coefficients with a mean value of 0.79. CONCLUSIONS: The proposed method is capable of obtaining the left ventricle volume biomarker in the end-diastole and offer an explanation of how it obtains the result in the form of a segmentation mask without the need of segmentation labels to train the algorithm, making it a potentially more trustworthy method for clinicians and a way to train neural networks more easily when segmentation labels are not readily available.
publishDate 2021
dc.date.none.fl_str_mv 2021
dc.type.none.fl_str_mv info:eu-repo/semantics/article
info:eu-repo/semantics/publishedVersion
format article
status_str publishedVersion
dc.identifier.none.fl_str_mv https://incliva.portalinvestigacion.com/publicaciones/15844
url https://incliva.portalinvestigacion.com/publicaciones/15844
dc.language.none.fl_str_mv Inglés
language_invalid_str_mv Inglés
dc.rights.none.fl_str_mv info:eu-repo/semantics/openAccess
eu_rights_str_mv openAccess
dc.publisher.none.fl_str_mv ELSEVIER IRELAND LTD
publisher.none.fl_str_mv ELSEVIER IRELAND LTD
dc.source.none.fl_str_mv Computer Methods and Programs in Biomedicine
ISSN: 01692607
ISSNe: 18727565
reponame:r-INCLIVA. Repositorio Institucional de Producción Científica de INCLIVA
instname:INCLIVA
instname_str INCLIVA
reponame_str r-INCLIVA. Repositorio Institucional de Producción Científica de INCLIVA
collection r-INCLIVA. Repositorio Institucional de Producción Científica de INCLIVA
repository.name.fl_str_mv
repository.mail.fl_str_mv
_version_ 1869411789663371264
score 15,811543

Similares en LA Referencia