Deep cascaded registration and weakly-supervised segmentation of fetal brain MRI
Deformable image registration is a cornerstone of many medical image analysis applications, particularly in the context of fetal brain magnetic resonance imaging (MRI), where precise registration is essential for studying the rapidly evolving fetal brain during pregnancy and potentially identifying...
Saved in:
| Main Authors: | , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-01-01
|
| Series: | Heliyon |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2405844024161798 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841526254924201984 |
|---|---|
| author | Valentin Comte Mireia Alenya Andrea Urru Judith Recober Ayako Nakaki Francesca Crovetto Oscar Camara Eduard Gratacós Elisenda Eixarch Fatima Crispi Gemma Piella Mario Ceresa Miguel A. González Ballester |
| author_facet | Valentin Comte Mireia Alenya Andrea Urru Judith Recober Ayako Nakaki Francesca Crovetto Oscar Camara Eduard Gratacós Elisenda Eixarch Fatima Crispi Gemma Piella Mario Ceresa Miguel A. González Ballester |
| author_sort | Valentin Comte |
| collection | DOAJ |
| description | Deformable image registration is a cornerstone of many medical image analysis applications, particularly in the context of fetal brain magnetic resonance imaging (MRI), where precise registration is essential for studying the rapidly evolving fetal brain during pregnancy and potentially identifying neurodevelopmental abnormalities. While deep learning has become the leading approach for medical image registration, traditional convolutional neural networks (CNNs) often fall short in capturing fine image details due to their bias toward low spatial frequencies. To address this challenge, we introduce a deep learning registration framework comprising multiple cascaded convolutional networks. These networks predict a series of incremental deformation fields that transform the moving image at various spatial frequency levels, ensuring accurate alignment with the fixed image. This multi-resolution approach allows for a more accurate and detailed registration process, capturing both coarse and fine image structures. Our method outperforms existing state-of-the-art techniques, including other multi-resolution strategies, by a substantial margin. Furthermore, we integrate our registration method into a multi-atlas segmentation pipeline and showcase its competitive performance compared to nnU-Net, achieved using only a small subset of annotated images as atlases. This approach is particularly valuable in the context of fetal brain MRI, where annotated datasets are limited. Our pipeline for registration and multi-atlas segmentation is publicly available at https://github.com/ValBcn/CasReg. |
| format | Article |
| id | doaj-art-29047c771abd44168826eef8a7d66440 |
| institution | Kabale University |
| issn | 2405-8440 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Heliyon |
| spelling | doaj-art-29047c771abd44168826eef8a7d664402025-01-17T04:49:40ZengElsevierHeliyon2405-84402025-01-01111e40148Deep cascaded registration and weakly-supervised segmentation of fetal brain MRIValentin Comte0Mireia Alenya1Andrea Urru2Judith Recober3Ayako Nakaki4Francesca Crovetto5Oscar Camara6Eduard Gratacós7Elisenda Eixarch8Fatima Crispi9Gemma Piella10Mario Ceresa11Miguel A. González Ballester12BCN MedTech, Department of Information and Communication Technologies, Universitat Pompeu Fabra, Barcelona, Spain; European Commission, Joint Research Centre (JRC), Geel, Belgium; Corresponding author. BCN MedTech, Department of Information and Communication Technologies, Universitat Pompeu Fabra, Barcelona, SpainBCN MedTech, Department of Information and Communication Technologies, Universitat Pompeu Fabra, Barcelona, SpainBCN MedTech, Department of Information and Communication Technologies, Universitat Pompeu Fabra, Barcelona, SpainBCN MedTech, Department of Information and Communication Technologies, Universitat Pompeu Fabra, Barcelona, SpainBCNatal, Fetal Medicine Research Center (Hospital Clínic and Hospital Sant Joan de Déu), University of Barcelona, Barcelona, Spain; Institut d’Investigacions Biomédiques August Pi i Sunyer (IDIBAPS), Barcelona, SpainBCNatal, Fetal Medicine Research Center (Hospital Clínic and Hospital Sant Joan de Déu), University of Barcelona, Barcelona, SpainBCN MedTech, Department of Information and Communication Technologies, Universitat Pompeu Fabra, Barcelona, SpainBCNatal, Fetal Medicine Research Center (Hospital Clínic and Hospital Sant Joan de Déu), University of Barcelona, Barcelona, Spain; Institut d’Investigacions Biomédiques August Pi i Sunyer (IDIBAPS), Barcelona, Spain; Centre for Biomedical Research on Rare Diseases (CIBERER), Barcelona, SpainBCNatal, Fetal Medicine Research Center (Hospital Clínic and Hospital Sant Joan de Déu), University of Barcelona, Barcelona, Spain; Institut d’Investigacions Biomédiques August Pi i Sunyer (IDIBAPS), Barcelona, Spain; Centre for Biomedical Research on Rare Diseases (CIBERER), Barcelona, SpainBCNatal, Fetal Medicine Research Center (Hospital Clínic and Hospital Sant Joan de Déu), University of Barcelona, Barcelona, Spain; Institut d’Investigacions Biomédiques August Pi i Sunyer (IDIBAPS), Barcelona, Spain; Centre for Biomedical Research on Rare Diseases (CIBERER), Barcelona, SpainBCN MedTech, Department of Information and Communication Technologies, Universitat Pompeu Fabra, Barcelona, SpainBCN MedTech, Department of Information and Communication Technologies, Universitat Pompeu Fabra, Barcelona, Spain; European Commission, Joint Research Centre (JRC), Ispra, ItalyBCN MedTech, Department of Information and Communication Technologies, Universitat Pompeu Fabra, Barcelona, Spain; ICREA, Barcelona, SpainDeformable image registration is a cornerstone of many medical image analysis applications, particularly in the context of fetal brain magnetic resonance imaging (MRI), where precise registration is essential for studying the rapidly evolving fetal brain during pregnancy and potentially identifying neurodevelopmental abnormalities. While deep learning has become the leading approach for medical image registration, traditional convolutional neural networks (CNNs) often fall short in capturing fine image details due to their bias toward low spatial frequencies. To address this challenge, we introduce a deep learning registration framework comprising multiple cascaded convolutional networks. These networks predict a series of incremental deformation fields that transform the moving image at various spatial frequency levels, ensuring accurate alignment with the fixed image. This multi-resolution approach allows for a more accurate and detailed registration process, capturing both coarse and fine image structures. Our method outperforms existing state-of-the-art techniques, including other multi-resolution strategies, by a substantial margin. Furthermore, we integrate our registration method into a multi-atlas segmentation pipeline and showcase its competitive performance compared to nnU-Net, achieved using only a small subset of annotated images as atlases. This approach is particularly valuable in the context of fetal brain MRI, where annotated datasets are limited. Our pipeline for registration and multi-atlas segmentation is publicly available at https://github.com/ValBcn/CasReg.http://www.sciencedirect.com/science/article/pii/S2405844024161798RegistrationSegmentationCascadeDeep learningFetal brain |
| spellingShingle | Valentin Comte Mireia Alenya Andrea Urru Judith Recober Ayako Nakaki Francesca Crovetto Oscar Camara Eduard Gratacós Elisenda Eixarch Fatima Crispi Gemma Piella Mario Ceresa Miguel A. González Ballester Deep cascaded registration and weakly-supervised segmentation of fetal brain MRI Heliyon Registration Segmentation Cascade Deep learning Fetal brain |
| title | Deep cascaded registration and weakly-supervised segmentation of fetal brain MRI |
| title_full | Deep cascaded registration and weakly-supervised segmentation of fetal brain MRI |
| title_fullStr | Deep cascaded registration and weakly-supervised segmentation of fetal brain MRI |
| title_full_unstemmed | Deep cascaded registration and weakly-supervised segmentation of fetal brain MRI |
| title_short | Deep cascaded registration and weakly-supervised segmentation of fetal brain MRI |
| title_sort | deep cascaded registration and weakly supervised segmentation of fetal brain mri |
| topic | Registration Segmentation Cascade Deep learning Fetal brain |
| url | http://www.sciencedirect.com/science/article/pii/S2405844024161798 |
| work_keys_str_mv | AT valentincomte deepcascadedregistrationandweaklysupervisedsegmentationoffetalbrainmri AT mireiaalenya deepcascadedregistrationandweaklysupervisedsegmentationoffetalbrainmri AT andreaurru deepcascadedregistrationandweaklysupervisedsegmentationoffetalbrainmri AT judithrecober deepcascadedregistrationandweaklysupervisedsegmentationoffetalbrainmri AT ayakonakaki deepcascadedregistrationandweaklysupervisedsegmentationoffetalbrainmri AT francescacrovetto deepcascadedregistrationandweaklysupervisedsegmentationoffetalbrainmri AT oscarcamara deepcascadedregistrationandweaklysupervisedsegmentationoffetalbrainmri AT eduardgratacos deepcascadedregistrationandweaklysupervisedsegmentationoffetalbrainmri AT elisendaeixarch deepcascadedregistrationandweaklysupervisedsegmentationoffetalbrainmri AT fatimacrispi deepcascadedregistrationandweaklysupervisedsegmentationoffetalbrainmri AT gemmapiella deepcascadedregistrationandweaklysupervisedsegmentationoffetalbrainmri AT marioceresa deepcascadedregistrationandweaklysupervisedsegmentationoffetalbrainmri AT miguelagonzalezballester deepcascadedregistrationandweaklysupervisedsegmentationoffetalbrainmri |