Recursive variational autoencoders for 3D blood vessel generative modeling

Autores
Feldman, Paula Adi; Fainstein, Miguel; Siless, Viviana; Delrieux, Claudio Augusto; Iarussi, Emmanuel
Año de publicación
2025
Idioma
inglés
Tipo de recurso
artículo
Estado
versión publicada
Descripción
Anatomical trees play an important role in clinical diagnosis and treatment planning. Yet, accurately representing these structures poses significant challenges owing to their intricate and varied topology and geometry. Most existing methods to synthesize vasculature are rule based, and despite providing some degree of control and variation in the structures produced, they fail to capture the diversity and complexity of actual anatomical data. We developed a Recursive variational Neural Network (RvNN) that fully exploits the hierarchical organization of the vessel and learns a low-dimensional manifold encoding branch connectivity along with geometry features describing the target surface. After training, the RvNN latent space can be sampled to generate new vessel geometries. By leveraging the power of generative neural networks, we generate 3D models of blood vessels that are both accurate and diverse, which is crucial for medical and surgical training, hemodynamic simulations, and many other purposes. These results closely resemble real data, achieving high similarity in vessel radii, length, and tortuosity across various datasets, including those with aneurysms. To the best of our knowledge, this work is the first to utilize this technique for synthesizing blood vessels.
Fil: Feldman, Paula Adi. Consejo Nacional de Investigaciones Científicas y Técnicas; Argentina. Universidad Torcuato Di Tella; Argentina
Fil: Fainstein, Miguel. Universidad Torcuato Di Tella; Argentina
Fil: Siless, Viviana. Universidad Torcuato Di Tella; Argentina
Fil: Delrieux, Claudio Augusto. Consejo Nacional de Investigaciones Científicas y Técnicas; Argentina. Universidad Nacional del Sur. Departamento de Ingeniería Eléctrica y de Computadoras; Argentina
Fil: Iarussi, Emmanuel. Consejo Nacional de Investigaciones Científicas y Técnicas; Argentina. Universidad Torcuato Di Tella; Argentina
Materia
VASCULAR 3D MODEL
GENERATIVE MODELING
NEURAL NETWORKS
VARIATIONAL AUTOENCODERS
Nivel de accesibilidad
acceso abierto
Condiciones de uso
https://creativecommons.org/licenses/by-nc-sa/2.5/ar/
Repositorio
CONICET Digital (CONICET)
Institución
Consejo Nacional de Investigaciones Científicas y Técnicas
OAI Identificador
oai:ri.conicet.gov.ar:11336/271369
Acceder
id CONICETDig_87a09947e9b518b9829b5dcbd4ffa59d
oai_identifier_str oai:ri.conicet.gov.ar:11336/271369
network_acronym_str CONICETDig
repository_id_str 3498
network_name_str CONICET Digital (CONICET)
spelling Recursive variational autoencoders for 3D blood vessel generative modelingFeldman, Paula AdiFainstein, MiguelSiless, VivianaDelrieux, Claudio AugustoIarussi, EmmanuelVASCULAR 3D MODELGENERATIVE MODELINGNEURAL NETWORKSVARIATIONAL AUTOENCODERShttps://purl.org/becyt/ford/1.2https://purl.org/becyt/ford/1Anatomical trees play an important role in clinical diagnosis and treatment planning. Yet, accurately representing these structures poses significant challenges owing to their intricate and varied topology and geometry. Most existing methods to synthesize vasculature are rule based, and despite providing some degree of control and variation in the structures produced, they fail to capture the diversity and complexity of actual anatomical data. We developed a Recursive variational Neural Network (RvNN) that fully exploits the hierarchical organization of the vessel and learns a low-dimensional manifold encoding branch connectivity along with geometry features describing the target surface. After training, the RvNN latent space can be sampled to generate new vessel geometries. By leveraging the power of generative neural networks, we generate 3D models of blood vessels that are both accurate and diverse, which is crucial for medical and surgical training, hemodynamic simulations, and many other purposes. These results closely resemble real data, achieving high similarity in vessel radii, length, and tortuosity across various datasets, including those with aneurysms. To the best of our knowledge, this work is the first to utilize this technique for synthesizing blood vessels.Fil: Feldman, Paula Adi. Consejo Nacional de Investigaciones Científicas y Técnicas; Argentina. Universidad Torcuato Di Tella; ArgentinaFil: Fainstein, Miguel. Universidad Torcuato Di Tella; ArgentinaFil: Siless, Viviana. Universidad Torcuato Di Tella; ArgentinaFil: Delrieux, Claudio Augusto. Consejo Nacional de Investigaciones Científicas y Técnicas; Argentina. Universidad Nacional del Sur. Departamento de Ingeniería Eléctrica y de Computadoras; ArgentinaFil: Iarussi, Emmanuel. Consejo Nacional de Investigaciones Científicas y Técnicas; Argentina. Universidad Torcuato Di Tella; ArgentinaElsevier Science2025-10info:eu-repo/semantics/articleinfo:eu-repo/semantics/publishedVersionhttp://purl.org/coar/resource_type/c_6501info:ar-repo/semantics/articuloapplication/pdfapplication/pdfapplication/pdfhttp://hdl.handle.net/11336/271369Feldman, Paula Adi; Fainstein, Miguel; Siless, Viviana; Delrieux, Claudio Augusto; Iarussi, Emmanuel; Recursive variational autoencoders for 3D blood vessel generative modeling; Elsevier Science; Medical Image Analysis; 105; 103703; 10-2025; 1-131361-8415CONICET DigitalCONICETenginfo:eu-repo/semantics/altIdentifier/url/https://linkinghub.elsevier.com/retrieve/pii/S1361841525002506info:eu-repo/semantics/altIdentifier/doi/10.1016/j.media.2025.103703info:eu-repo/semantics/altIdentifier/url/https://arxiv.org/abs/2506.14914v1info:eu-repo/semantics/openAccesshttps://creativecommons.org/licenses/by-nc-sa/2.5/ar/reponame:CONICET Digital (CONICET)instname:Consejo Nacional de Investigaciones Científicas y Técnicas2025-09-29T10:06:21Zoai:ri.conicet.gov.ar:11336/271369instacron:CONICETInstitucionalhttp://ri.conicet.gov.ar/Organismo científico-tecnológicoNo correspondehttp://ri.conicet.gov.ar/oai/requestdasensio@conicet.gov.ar; lcarlino@conicet.gov.arArgentinaNo correspondeNo correspondeNo correspondeopendoar:34982025-09-29 10:06:21.908CONICET Digital (CONICET) - Consejo Nacional de Investigaciones Científicas y Técnicasfalse
dc.title.none.fl_str_mv Recursive variational autoencoders for 3D blood vessel generative modeling
title Recursive variational autoencoders for 3D blood vessel generative modeling
spellingShingle Recursive variational autoencoders for 3D blood vessel generative modeling
Feldman, Paula Adi
VASCULAR 3D MODEL
GENERATIVE MODELING
NEURAL NETWORKS
VARIATIONAL AUTOENCODERS
title_short Recursive variational autoencoders for 3D blood vessel generative modeling
title_full Recursive variational autoencoders for 3D blood vessel generative modeling
title_fullStr Recursive variational autoencoders for 3D blood vessel generative modeling
title_full_unstemmed Recursive variational autoencoders for 3D blood vessel generative modeling
title_sort Recursive variational autoencoders for 3D blood vessel generative modeling
dc.creator.none.fl_str_mv Feldman, Paula Adi
Fainstein, Miguel
Siless, Viviana
Delrieux, Claudio Augusto
Iarussi, Emmanuel
author Feldman, Paula Adi
author_facet Feldman, Paula Adi
Fainstein, Miguel
Siless, Viviana
Delrieux, Claudio Augusto
Iarussi, Emmanuel
author_role author
author2 Fainstein, Miguel
Siless, Viviana
Delrieux, Claudio Augusto
Iarussi, Emmanuel
author2_role author
author
author
author
dc.subject.none.fl_str_mv VASCULAR 3D MODEL
GENERATIVE MODELING
NEURAL NETWORKS
VARIATIONAL AUTOENCODERS
topic VASCULAR 3D MODEL
GENERATIVE MODELING
NEURAL NETWORKS
VARIATIONAL AUTOENCODERS
purl_subject.fl_str_mv https://purl.org/becyt/ford/1.2
https://purl.org/becyt/ford/1
dc.description.none.fl_txt_mv Anatomical trees play an important role in clinical diagnosis and treatment planning. Yet, accurately representing these structures poses significant challenges owing to their intricate and varied topology and geometry. Most existing methods to synthesize vasculature are rule based, and despite providing some degree of control and variation in the structures produced, they fail to capture the diversity and complexity of actual anatomical data. We developed a Recursive variational Neural Network (RvNN) that fully exploits the hierarchical organization of the vessel and learns a low-dimensional manifold encoding branch connectivity along with geometry features describing the target surface. After training, the RvNN latent space can be sampled to generate new vessel geometries. By leveraging the power of generative neural networks, we generate 3D models of blood vessels that are both accurate and diverse, which is crucial for medical and surgical training, hemodynamic simulations, and many other purposes. These results closely resemble real data, achieving high similarity in vessel radii, length, and tortuosity across various datasets, including those with aneurysms. To the best of our knowledge, this work is the first to utilize this technique for synthesizing blood vessels.
Fil: Feldman, Paula Adi. Consejo Nacional de Investigaciones Científicas y Técnicas; Argentina. Universidad Torcuato Di Tella; Argentina
Fil: Fainstein, Miguel. Universidad Torcuato Di Tella; Argentina
Fil: Siless, Viviana. Universidad Torcuato Di Tella; Argentina
Fil: Delrieux, Claudio Augusto. Consejo Nacional de Investigaciones Científicas y Técnicas; Argentina. Universidad Nacional del Sur. Departamento de Ingeniería Eléctrica y de Computadoras; Argentina
Fil: Iarussi, Emmanuel. Consejo Nacional de Investigaciones Científicas y Técnicas; Argentina. Universidad Torcuato Di Tella; Argentina
description Anatomical trees play an important role in clinical diagnosis and treatment planning. Yet, accurately representing these structures poses significant challenges owing to their intricate and varied topology and geometry. Most existing methods to synthesize vasculature are rule based, and despite providing some degree of control and variation in the structures produced, they fail to capture the diversity and complexity of actual anatomical data. We developed a Recursive variational Neural Network (RvNN) that fully exploits the hierarchical organization of the vessel and learns a low-dimensional manifold encoding branch connectivity along with geometry features describing the target surface. After training, the RvNN latent space can be sampled to generate new vessel geometries. By leveraging the power of generative neural networks, we generate 3D models of blood vessels that are both accurate and diverse, which is crucial for medical and surgical training, hemodynamic simulations, and many other purposes. These results closely resemble real data, achieving high similarity in vessel radii, length, and tortuosity across various datasets, including those with aneurysms. To the best of our knowledge, this work is the first to utilize this technique for synthesizing blood vessels.
publishDate 2025
dc.date.none.fl_str_mv 2025-10
dc.type.none.fl_str_mv info:eu-repo/semantics/article
info:eu-repo/semantics/publishedVersion
http://purl.org/coar/resource_type/c_6501
info:ar-repo/semantics/articulo
format article
status_str publishedVersion
dc.identifier.none.fl_str_mv http://hdl.handle.net/11336/271369
Feldman, Paula Adi; Fainstein, Miguel; Siless, Viviana; Delrieux, Claudio Augusto; Iarussi, Emmanuel; Recursive variational autoencoders for 3D blood vessel generative modeling; Elsevier Science; Medical Image Analysis; 105; 103703; 10-2025; 1-13
1361-8415
CONICET Digital
CONICET
url http://hdl.handle.net/11336/271369
identifier_str_mv Feldman, Paula Adi; Fainstein, Miguel; Siless, Viviana; Delrieux, Claudio Augusto; Iarussi, Emmanuel; Recursive variational autoencoders for 3D blood vessel generative modeling; Elsevier Science; Medical Image Analysis; 105; 103703; 10-2025; 1-13
1361-8415
CONICET Digital
CONICET
dc.language.none.fl_str_mv eng
language eng
dc.relation.none.fl_str_mv info:eu-repo/semantics/altIdentifier/url/https://linkinghub.elsevier.com/retrieve/pii/S1361841525002506
info:eu-repo/semantics/altIdentifier/doi/10.1016/j.media.2025.103703
info:eu-repo/semantics/altIdentifier/url/https://arxiv.org/abs/2506.14914v1
dc.rights.none.fl_str_mv info:eu-repo/semantics/openAccess
https://creativecommons.org/licenses/by-nc-sa/2.5/ar/
eu_rights_str_mv openAccess
rights_invalid_str_mv https://creativecommons.org/licenses/by-nc-sa/2.5/ar/
dc.format.none.fl_str_mv application/pdf
application/pdf
application/pdf
dc.publisher.none.fl_str_mv Elsevier Science
publisher.none.fl_str_mv Elsevier Science
dc.source.none.fl_str_mv reponame:CONICET Digital (CONICET)
instname:Consejo Nacional de Investigaciones Científicas y Técnicas
reponame_str CONICET Digital (CONICET)
collection CONICET Digital (CONICET)
instname_str Consejo Nacional de Investigaciones Científicas y Técnicas
repository.name.fl_str_mv CONICET Digital (CONICET) - Consejo Nacional de Investigaciones Científicas y Técnicas
repository.mail.fl_str_mv dasensio@conicet.gov.ar; lcarlino@conicet.gov.ar
_version_ 1844613911213506560
score 13.070432

Publicaciones similares