Cargando…

Automated olfactory bulb segmentation on high resolutional T2-weighted MRI

The neuroimage analysis community has neglected the automated segmentation of the olfactory bulb (OB) despite its crucial role in olfactory function. The lack of an automatic processing method for the OB can be explained by its challenging properties (small size, location, and poor visibility on tra...

Descripción completa

Detalles Bibliográficos
Autores principales: Estrada, Santiago, Lu, Ran, Diers, Kersten, Zeng, Weiyi, Ehses, Philipp, Stöcker, Tony, Breteler, Monique M. B, Reuter, Martin
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Academic Press 2021
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8473894/
https://www.ncbi.nlm.nih.gov/pubmed/34389442
http://dx.doi.org/10.1016/j.neuroimage.2021.118464
_version_ 1784575097031360512
author Estrada, Santiago
Lu, Ran
Diers, Kersten
Zeng, Weiyi
Ehses, Philipp
Stöcker, Tony
Breteler, Monique M. B
Reuter, Martin
author_facet Estrada, Santiago
Lu, Ran
Diers, Kersten
Zeng, Weiyi
Ehses, Philipp
Stöcker, Tony
Breteler, Monique M. B
Reuter, Martin
author_sort Estrada, Santiago
collection PubMed
description The neuroimage analysis community has neglected the automated segmentation of the olfactory bulb (OB) despite its crucial role in olfactory function. The lack of an automatic processing method for the OB can be explained by its challenging properties (small size, location, and poor visibility on traditional MRI scans). Nonetheless, recent advances in MRI acquisition techniques and resolution have allowed raters to generate more reliable manual annotations. Furthermore, the high accuracy of deep learning methods for solving semantic segmentation problems provides us with an option to reliably assess even small structures. In this work, we introduce a novel, fast, and fully automated deep learning pipeline to accurately segment OB tissue on sub-millimeter T2-weighted (T2w) whole-brain MR images. To this end, we designed a three-stage pipeline: (1) Localization of a region containing both OBs using FastSurferCNN, (2) Segmentation of OB tissue within the localized region through four independent AttFastSurferCNN - a novel deep learning architecture with a self-attention mechanism to improve modeling of contextual information, and (3) Ensemble of the predicted label maps. For this work, both OBs were manually annotated in a total of 620 T2w images for training (n=357) and testing. The OB pipeline exhibits high performance in terms of boundary delineation, OB localization, and volume estimation across a wide range of ages in 203 participants of the Rhineland Study (Dice Score (Dice): 0.852, Volume Similarity (VS): 0.910, and Average Hausdorff Distance (AVD): 0.215  [Formula: see text]). Moreover, it also generalizes to scans of an independent dataset never encountered during training, the Human Connectome Project (HCP), with different acquisition parameters and demographics, evaluated in 30 cases at the native 0.7  [Formula: see text] HCP resolution (Dice: 0.738, VS: 0.790, and AVD: 0.340  [Formula: see text]), and the default 0.8  [Formula: see text] pipeline resolution (Dice: 0.782, VS: 0.858, and AVD: 0.268  [Formula: see text]). We extensively validated our pipeline not only with respect to segmentation accuracy but also to known OB volume effects, where it can sensitively replicate age effects ([Formula: see text] , [Formula: see text]). Furthermore, our method can analyze a 3D volume in less than a minute (GPU) in an end-to-end fashion, providing a validated, efficient, and scalable solution for automatically assessing OB volumes.
format Online
Article
Text
id pubmed-8473894
institution National Center for Biotechnology Information
language English
publishDate 2021
publisher Academic Press
record_format MEDLINE/PubMed
spelling pubmed-84738942021-11-15 Automated olfactory bulb segmentation on high resolutional T2-weighted MRI Estrada, Santiago Lu, Ran Diers, Kersten Zeng, Weiyi Ehses, Philipp Stöcker, Tony Breteler, Monique M. B Reuter, Martin Neuroimage Article The neuroimage analysis community has neglected the automated segmentation of the olfactory bulb (OB) despite its crucial role in olfactory function. The lack of an automatic processing method for the OB can be explained by its challenging properties (small size, location, and poor visibility on traditional MRI scans). Nonetheless, recent advances in MRI acquisition techniques and resolution have allowed raters to generate more reliable manual annotations. Furthermore, the high accuracy of deep learning methods for solving semantic segmentation problems provides us with an option to reliably assess even small structures. In this work, we introduce a novel, fast, and fully automated deep learning pipeline to accurately segment OB tissue on sub-millimeter T2-weighted (T2w) whole-brain MR images. To this end, we designed a three-stage pipeline: (1) Localization of a region containing both OBs using FastSurferCNN, (2) Segmentation of OB tissue within the localized region through four independent AttFastSurferCNN - a novel deep learning architecture with a self-attention mechanism to improve modeling of contextual information, and (3) Ensemble of the predicted label maps. For this work, both OBs were manually annotated in a total of 620 T2w images for training (n=357) and testing. The OB pipeline exhibits high performance in terms of boundary delineation, OB localization, and volume estimation across a wide range of ages in 203 participants of the Rhineland Study (Dice Score (Dice): 0.852, Volume Similarity (VS): 0.910, and Average Hausdorff Distance (AVD): 0.215  [Formula: see text]). Moreover, it also generalizes to scans of an independent dataset never encountered during training, the Human Connectome Project (HCP), with different acquisition parameters and demographics, evaluated in 30 cases at the native 0.7  [Formula: see text] HCP resolution (Dice: 0.738, VS: 0.790, and AVD: 0.340  [Formula: see text]), and the default 0.8  [Formula: see text] pipeline resolution (Dice: 0.782, VS: 0.858, and AVD: 0.268  [Formula: see text]). We extensively validated our pipeline not only with respect to segmentation accuracy but also to known OB volume effects, where it can sensitively replicate age effects ([Formula: see text] , [Formula: see text]). Furthermore, our method can analyze a 3D volume in less than a minute (GPU) in an end-to-end fashion, providing a validated, efficient, and scalable solution for automatically assessing OB volumes. Academic Press 2021-11-15 /pmc/articles/PMC8473894/ /pubmed/34389442 http://dx.doi.org/10.1016/j.neuroimage.2021.118464 Text en © 2021 The Author(s) https://creativecommons.org/licenses/by/4.0/This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
spellingShingle Article
Estrada, Santiago
Lu, Ran
Diers, Kersten
Zeng, Weiyi
Ehses, Philipp
Stöcker, Tony
Breteler, Monique M. B
Reuter, Martin
Automated olfactory bulb segmentation on high resolutional T2-weighted MRI
title Automated olfactory bulb segmentation on high resolutional T2-weighted MRI
title_full Automated olfactory bulb segmentation on high resolutional T2-weighted MRI
title_fullStr Automated olfactory bulb segmentation on high resolutional T2-weighted MRI
title_full_unstemmed Automated olfactory bulb segmentation on high resolutional T2-weighted MRI
title_short Automated olfactory bulb segmentation on high resolutional T2-weighted MRI
title_sort automated olfactory bulb segmentation on high resolutional t2-weighted mri
topic Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8473894/
https://www.ncbi.nlm.nih.gov/pubmed/34389442
http://dx.doi.org/10.1016/j.neuroimage.2021.118464
work_keys_str_mv AT estradasantiago automatedolfactorybulbsegmentationonhighresolutionalt2weightedmri
AT luran automatedolfactorybulbsegmentationonhighresolutionalt2weightedmri
AT dierskersten automatedolfactorybulbsegmentationonhighresolutionalt2weightedmri
AT zengweiyi automatedolfactorybulbsegmentationonhighresolutionalt2weightedmri
AT ehsesphilipp automatedolfactorybulbsegmentationonhighresolutionalt2weightedmri
AT stockertony automatedolfactorybulbsegmentationonhighresolutionalt2weightedmri
AT bretelermoniquemb automatedolfactorybulbsegmentationonhighresolutionalt2weightedmri
AT reutermartin automatedolfactorybulbsegmentationonhighresolutionalt2weightedmri