G-CNN: Adaptive Geometric Convolutional Neural Networks for MRI-Based Skull Stripping

Yifan Li; Chao Li; Yiran Wei; Stephen J. Price; Carola-Bibiane Schoenlieb; Xi Chen

doi:10.1007/978-3-031-45087-7_3

G-CNN: Adaptive Geometric Convolutional Neural Networks for MRI-Based Skull Stripping

Yifan Li, Chao Li, Yiran Wei, Stephen J. Price, Carola-Bibiane Schoenlieb, Xi Chen

Research output: Chapter or section in a book/report/conference proceeding › Chapter in a published conference proceeding

Abstract

Skull stripping in MRI-based brain imaging involves extraction of brain regions from raw images. While some Convolutional Neural Nets (CNNs) models have been successful in automating this process, the reliance on local textures can negatively impact model performance in the presence of pathological conditions such as brain tumors. This study presents a novel yet practical approach to offer supplementary texture-invariant spatial information of the brain as a geometric prior to enhance shape representations to further enable informed segmentation decisions. Our numerical results demonstrate that the new method outperforms competing algorithms in skull stripping tasks for both healthy and pathological inputs. These findings underscore the potential of incorporating geometric information in deep learning models to enhance the accuracy of brain segmentation.

Original language	English
Title of host publication	International MICCAI Workshop 2023
Subtitle of host publication	The 2nd Workshop on Computational Mathematics Modeling in Cancer Analysis
Place of Publication	Cham, Switzerland
Publisher	Springer, Cham
Pages	21-30
Number of pages	10
ISBN (Electronic)	9783031450877
ISBN (Print)	9783031450860
DOIs	https://doi.org/10.1007/978-3-031-45087-7_3
Publication status	Published - 8 Oct 2023
Event	2nd Workshop on Computational Mathematics Modeling in Cancer Analysis - Duration: 8 Oct 2023 → …

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	14243
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	2nd Workshop on Computational Mathematics Modeling in Cancer Analysis
Period	8/10/23 → …

Keywords

Skull stripping
Geometric information
Deep learning
Convolutional neural network (CNN)

Access to Document

10.1007/978-3-031-45087-7_3

Cite this

Li, Y., Li, C., Wei, Y., Price, S. J., Schoenlieb, C.-B., & Chen, X. (2023). G-CNN: Adaptive Geometric Convolutional Neural Networks for MRI-Based Skull Stripping. In International MICCAI Workshop 2023: The 2nd Workshop on Computational Mathematics Modeling in Cancer Analysis (pp. 21-30). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 14243). Springer, Cham. https://doi.org/10.1007/978-3-031-45087-7_3

G-CNN: Adaptive Geometric Convolutional Neural Networks for MRI-Based Skull Stripping. / Li, Yifan; Li, Chao; Wei, Yiran et al.
International MICCAI Workshop 2023: The 2nd Workshop on Computational Mathematics Modeling in Cancer Analysis. Cham, Switzerland: Springer, Cham, 2023. p. 21-30 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 14243).

Research output: Chapter or section in a book/report/conference proceeding › Chapter in a published conference proceeding

Li, Y, Li, C, Wei, Y, Price, SJ, Schoenlieb, C-B & Chen, X 2023, G-CNN: Adaptive Geometric Convolutional Neural Networks for MRI-Based Skull Stripping. in International MICCAI Workshop 2023: The 2nd Workshop on Computational Mathematics Modeling in Cancer Analysis. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 14243, Springer, Cham, Cham, Switzerland, pp. 21-30, 2nd Workshop on Computational Mathematics Modeling in Cancer Analysis, 8/10/23. https://doi.org/10.1007/978-3-031-45087-7_3

Li Y, Li C, Wei Y, Price SJ, Schoenlieb CB, Chen X. G-CNN: Adaptive Geometric Convolutional Neural Networks for MRI-Based Skull Stripping. In International MICCAI Workshop 2023: The 2nd Workshop on Computational Mathematics Modeling in Cancer Analysis. Cham, Switzerland: Springer, Cham. 2023. p. 21-30. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-031-45087-7_3

Li, Yifan ; Li, Chao ; Wei, Yiran et al. / G-CNN: Adaptive Geometric Convolutional Neural Networks for MRI-Based Skull Stripping. International MICCAI Workshop 2023: The 2nd Workshop on Computational Mathematics Modeling in Cancer Analysis. Cham, Switzerland : Springer, Cham, 2023. pp. 21-30 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{93e238e0593040388a076ef713270854,

title = "G-CNN: Adaptive Geometric Convolutional Neural Networks for MRI-Based Skull Stripping",

abstract = "Skull stripping in MRI-based brain imaging involves extraction of brain regions from raw images. While some Convolutional Neural Nets (CNNs) models have been successful in automating this process, the reliance on local textures can negatively impact model performance in the presence of pathological conditions such as brain tumors. This study presents a novel yet practical approach to offer supplementary texture-invariant spatial information of the brain as a geometric prior to enhance shape representations to further enable informed segmentation decisions. Our numerical results demonstrate that the new method outperforms competing algorithms in skull stripping tasks for both healthy and pathological inputs. These findings underscore the potential of incorporating geometric information in deep learning models to enhance the accuracy of brain segmentation.",

keywords = "Skull stripping, Geometric information, Deep learning, Convolutional neural network (CNN)",

author = "Yifan Li and Chao Li and Yiran Wei and Price, {Stephen J.} and Carola-Bibiane Schoenlieb and Xi Chen",

year = "2023",

month = oct,

day = "8",

doi = "10.1007/978-3-031-45087-7_3",

language = "English",

isbn = "9783031450860",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer, Cham",

pages = "21--30",

booktitle = "International MICCAI Workshop 2023",

note = " 2nd Workshop on Computational Mathematics Modeling in Cancer Analysis ; Conference date: 08-10-2023",

}

TY - GEN

T1 - G-CNN: Adaptive Geometric Convolutional Neural Networks for MRI-Based Skull Stripping

AU - Li, Yifan

AU - Li, Chao

AU - Wei, Yiran

AU - Price, Stephen J.

AU - Schoenlieb, Carola-Bibiane

AU - Chen, Xi

PY - 2023/10/8

Y1 - 2023/10/8

N2 - Skull stripping in MRI-based brain imaging involves extraction of brain regions from raw images. While some Convolutional Neural Nets (CNNs) models have been successful in automating this process, the reliance on local textures can negatively impact model performance in the presence of pathological conditions such as brain tumors. This study presents a novel yet practical approach to offer supplementary texture-invariant spatial information of the brain as a geometric prior to enhance shape representations to further enable informed segmentation decisions. Our numerical results demonstrate that the new method outperforms competing algorithms in skull stripping tasks for both healthy and pathological inputs. These findings underscore the potential of incorporating geometric information in deep learning models to enhance the accuracy of brain segmentation.

AB - Skull stripping in MRI-based brain imaging involves extraction of brain regions from raw images. While some Convolutional Neural Nets (CNNs) models have been successful in automating this process, the reliance on local textures can negatively impact model performance in the presence of pathological conditions such as brain tumors. This study presents a novel yet practical approach to offer supplementary texture-invariant spatial information of the brain as a geometric prior to enhance shape representations to further enable informed segmentation decisions. Our numerical results demonstrate that the new method outperforms competing algorithms in skull stripping tasks for both healthy and pathological inputs. These findings underscore the potential of incorporating geometric information in deep learning models to enhance the accuracy of brain segmentation.

KW - Skull stripping

KW - Geometric information

KW - Deep learning

KW - Convolutional neural network (CNN)

U2 - 10.1007/978-3-031-45087-7_3

DO - 10.1007/978-3-031-45087-7_3

M3 - Chapter in a published conference proceeding

SN - 9783031450860

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 21

EP - 30

BT - International MICCAI Workshop 2023

PB - Springer, Cham

CY - Cham, Switzerland

T2 - 2nd Workshop on Computational Mathematics Modeling in Cancer Analysis

Y2 - 8 October 2023

ER -

G-CNN: Adaptive Geometric Convolutional Neural Networks for MRI-Based Skull Stripping

Abstract

Publication series

Conference

Keywords

Access to Document

Fingerprint

Cite this