Comparison of Two-Dimensional- and Three-Dimensional-Based U-Net Architectures for Brain Tissue Classification in One-Dimensional Brain CT
Brain tissue segmentation plays a crucial role in feature extraction, volumetric quantification, and morphometric analysis of brain scans. For the assessment of brain structure and integrity, CT is a non-invasive, cheaper, faster, and more widely available modality than MRI. However, the clinical ap...
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2022-01-01
|
| Series: | Frontiers in Computational Neuroscience |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fncom.2021.785244/full |
| _version_ | 1818943366808731648 |
|---|---|
| author | Meera Srikrishna Meera Srikrishna Rolf A. Heckemann Joana B. Pereira Joana B. Pereira Giovanni Volpe Anna Zettergren Silke Kern Silke Kern Eric Westman Ingmar Skoog Ingmar Skoog Michael Schöll Michael Schöll Michael Schöll Michael Schöll |
| author_facet | Meera Srikrishna Meera Srikrishna Rolf A. Heckemann Joana B. Pereira Joana B. Pereira Giovanni Volpe Anna Zettergren Silke Kern Silke Kern Eric Westman Ingmar Skoog Ingmar Skoog Michael Schöll Michael Schöll Michael Schöll Michael Schöll |
| author_sort | Meera Srikrishna |
| collection | DOAJ |
| description | Brain tissue segmentation plays a crucial role in feature extraction, volumetric quantification, and morphometric analysis of brain scans. For the assessment of brain structure and integrity, CT is a non-invasive, cheaper, faster, and more widely available modality than MRI. However, the clinical application of CT is mostly limited to the visual assessment of brain integrity and exclusion of copathologies. We have previously developed two-dimensional (2D) deep learning-based segmentation networks that successfully classified brain tissue in head CT. Recently, deep learning-based MRI segmentation models successfully use patch-based three-dimensional (3D) segmentation networks. In this study, we aimed to develop patch-based 3D segmentation networks for CT brain tissue classification. Furthermore, we aimed to compare the performance of 2D- and 3D-based segmentation networks to perform brain tissue classification in anisotropic CT scans. For this purpose, we developed 2D and 3D U-Net-based deep learning models that were trained and validated on MR-derived segmentations from scans of 744 participants of the Gothenburg H70 Cohort with both CT and T1-weighted MRI scans acquired timely close to each other. Segmentation performance of both 2D and 3D models was evaluated on 234 unseen datasets using measures of distance, spatial similarity, and tissue volume. Single-task slice-wise processed 2D U-Nets performed better than multitask patch-based 3D U-Nets in CT brain tissue classification. These findings provide support to the use of 2D U-Nets to segment brain tissue in one-dimensional (1D) CT. This could increase the application of CT to detect brain abnormalities in clinical settings. |
| first_indexed | 2024-12-20T07:26:11Z |
| format | Article |
| id | doaj.art-f332efdd9abb44f6856bc369f0ee20ee |
| institution | Directory Open Access Journal |
| issn | 1662-5188 |
| language | English |
| last_indexed | 2024-12-20T07:26:11Z |
| publishDate | 2022-01-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Computational Neuroscience |
| spelling | doaj.art-f332efdd9abb44f6856bc369f0ee20ee2022-12-21T19:48:32ZengFrontiers Media S.A.Frontiers in Computational Neuroscience1662-51882022-01-011510.3389/fncom.2021.785244785244Comparison of Two-Dimensional- and Three-Dimensional-Based U-Net Architectures for Brain Tissue Classification in One-Dimensional Brain CTMeera Srikrishna0Meera Srikrishna1Rolf A. Heckemann2Joana B. Pereira3Joana B. Pereira4Giovanni Volpe5Anna Zettergren6Silke Kern7Silke Kern8Eric Westman9Ingmar Skoog10Ingmar Skoog11Michael Schöll12Michael Schöll13Michael Schöll14Michael Schöll15Wallenberg Centre for Molecular and Translational Medicine, University of Gothenburg, Gothenburg, SwedenDepartment of Psychiatry and Neurochemistry, Institute of Physiology and Neuroscience, University of Gothenburg, Gothenburg, SwedenDepartment of Medical Radiation Sciences, Institute of Clinical Sciences, Sahlgrenska Academy, Gothenburg, SwedenDivision of Clinical Geriatrics, Department of Neurobiology, Care Sciences and Society, Karolinska Institutet, Stockholm, SwedenMemory Research Unit, Department of Clinical Sciences, Malmö Lund University, Mälmo, SwedenDepartment of Physics, University of Gothenburg, Gothenburg, SwedenNeuropsychiatric Epidemiology, Institute of Neuroscience and Physiology, Sahlgrenska Academy, Centre for Ageing and Health (AgeCap), University of Gothenburg, Gothenburg, SwedenNeuropsychiatric Epidemiology, Institute of Neuroscience and Physiology, Sahlgrenska Academy, Centre for Ageing and Health (AgeCap), University of Gothenburg, Gothenburg, SwedenRegion Västra Götaland, Sahlgrenska University Hospital, Psychiatry, Cognition and Old Age Psychiatry Clinic, Gothenburg, SwedenDivision of Clinical Geriatrics, Department of Neurobiology, Care Sciences and Society, Karolinska Institutet, Stockholm, SwedenNeuropsychiatric Epidemiology, Institute of Neuroscience and Physiology, Sahlgrenska Academy, Centre for Ageing and Health (AgeCap), University of Gothenburg, Gothenburg, SwedenRegion Västra Götaland, Sahlgrenska University Hospital, Psychiatry, Cognition and Old Age Psychiatry Clinic, Gothenburg, SwedenWallenberg Centre for Molecular and Translational Medicine, University of Gothenburg, Gothenburg, SwedenDepartment of Psychiatry and Neurochemistry, Institute of Physiology and Neuroscience, University of Gothenburg, Gothenburg, SwedenDementia Research Centre, Institute of Neurology, University College London, London, United Kingdom0Department of Clinical Physiology, Sahlgrenska University Hospital, Gothenburg, SwedenBrain tissue segmentation plays a crucial role in feature extraction, volumetric quantification, and morphometric analysis of brain scans. For the assessment of brain structure and integrity, CT is a non-invasive, cheaper, faster, and more widely available modality than MRI. However, the clinical application of CT is mostly limited to the visual assessment of brain integrity and exclusion of copathologies. We have previously developed two-dimensional (2D) deep learning-based segmentation networks that successfully classified brain tissue in head CT. Recently, deep learning-based MRI segmentation models successfully use patch-based three-dimensional (3D) segmentation networks. In this study, we aimed to develop patch-based 3D segmentation networks for CT brain tissue classification. Furthermore, we aimed to compare the performance of 2D- and 3D-based segmentation networks to perform brain tissue classification in anisotropic CT scans. For this purpose, we developed 2D and 3D U-Net-based deep learning models that were trained and validated on MR-derived segmentations from scans of 744 participants of the Gothenburg H70 Cohort with both CT and T1-weighted MRI scans acquired timely close to each other. Segmentation performance of both 2D and 3D models was evaluated on 234 unseen datasets using measures of distance, spatial similarity, and tissue volume. Single-task slice-wise processed 2D U-Nets performed better than multitask patch-based 3D U-Nets in CT brain tissue classification. These findings provide support to the use of 2D U-Nets to segment brain tissue in one-dimensional (1D) CT. This could increase the application of CT to detect brain abnormalities in clinical settings.https://www.frontiersin.org/articles/10.3389/fncom.2021.785244/fullbrain image segmentationCTMRIdeep learningconvolutional neural networks |
| spellingShingle | Meera Srikrishna Meera Srikrishna Rolf A. Heckemann Joana B. Pereira Joana B. Pereira Giovanni Volpe Anna Zettergren Silke Kern Silke Kern Eric Westman Ingmar Skoog Ingmar Skoog Michael Schöll Michael Schöll Michael Schöll Michael Schöll Comparison of Two-Dimensional- and Three-Dimensional-Based U-Net Architectures for Brain Tissue Classification in One-Dimensional Brain CT Frontiers in Computational Neuroscience brain image segmentation CT MRI deep learning convolutional neural networks |
| title | Comparison of Two-Dimensional- and Three-Dimensional-Based U-Net Architectures for Brain Tissue Classification in One-Dimensional Brain CT |
| title_full | Comparison of Two-Dimensional- and Three-Dimensional-Based U-Net Architectures for Brain Tissue Classification in One-Dimensional Brain CT |
| title_fullStr | Comparison of Two-Dimensional- and Three-Dimensional-Based U-Net Architectures for Brain Tissue Classification in One-Dimensional Brain CT |
| title_full_unstemmed | Comparison of Two-Dimensional- and Three-Dimensional-Based U-Net Architectures for Brain Tissue Classification in One-Dimensional Brain CT |
| title_short | Comparison of Two-Dimensional- and Three-Dimensional-Based U-Net Architectures for Brain Tissue Classification in One-Dimensional Brain CT |
| title_sort | comparison of two dimensional and three dimensional based u net architectures for brain tissue classification in one dimensional brain ct |
| topic | brain image segmentation CT MRI deep learning convolutional neural networks |
| url | https://www.frontiersin.org/articles/10.3389/fncom.2021.785244/full |
| work_keys_str_mv | AT meerasrikrishna comparisonoftwodimensionalandthreedimensionalbasedunetarchitecturesforbraintissueclassificationinonedimensionalbrainct AT meerasrikrishna comparisonoftwodimensionalandthreedimensionalbasedunetarchitecturesforbraintissueclassificationinonedimensionalbrainct AT rolfaheckemann comparisonoftwodimensionalandthreedimensionalbasedunetarchitecturesforbraintissueclassificationinonedimensionalbrainct AT joanabpereira comparisonoftwodimensionalandthreedimensionalbasedunetarchitecturesforbraintissueclassificationinonedimensionalbrainct AT joanabpereira comparisonoftwodimensionalandthreedimensionalbasedunetarchitecturesforbraintissueclassificationinonedimensionalbrainct AT giovannivolpe comparisonoftwodimensionalandthreedimensionalbasedunetarchitecturesforbraintissueclassificationinonedimensionalbrainct AT annazettergren comparisonoftwodimensionalandthreedimensionalbasedunetarchitecturesforbraintissueclassificationinonedimensionalbrainct AT silkekern comparisonoftwodimensionalandthreedimensionalbasedunetarchitecturesforbraintissueclassificationinonedimensionalbrainct AT silkekern comparisonoftwodimensionalandthreedimensionalbasedunetarchitecturesforbraintissueclassificationinonedimensionalbrainct AT ericwestman comparisonoftwodimensionalandthreedimensionalbasedunetarchitecturesforbraintissueclassificationinonedimensionalbrainct AT ingmarskoog comparisonoftwodimensionalandthreedimensionalbasedunetarchitecturesforbraintissueclassificationinonedimensionalbrainct AT ingmarskoog comparisonoftwodimensionalandthreedimensionalbasedunetarchitecturesforbraintissueclassificationinonedimensionalbrainct AT michaelscholl comparisonoftwodimensionalandthreedimensionalbasedunetarchitecturesforbraintissueclassificationinonedimensionalbrainct AT michaelscholl comparisonoftwodimensionalandthreedimensionalbasedunetarchitecturesforbraintissueclassificationinonedimensionalbrainct AT michaelscholl comparisonoftwodimensionalandthreedimensionalbasedunetarchitecturesforbraintissueclassificationinonedimensionalbrainct AT michaelscholl comparisonoftwodimensionalandthreedimensionalbasedunetarchitecturesforbraintissueclassificationinonedimensionalbrainct |