A robust volumetric Transformer for accurate 3D tumor segmentation

Himashi Peiris; Munawar Hayat; Zhaolin Chen; Gary Egan; Mehrtash Harandi

doi:10.1007/978-3-031-16443-9_16

A robust volumetric Transformer for accurate 3D tumor segmentation

Himashi Peiris, Munawar Hayat, Zhaolin Chen, Gary Egan, Mehrtash Harandi

Research output: Chapter in Book/Report/Conference proceeding › Conference Paper › Research › peer-review

143 Citations (Scopus)

Abstract

We propose a Transformer architecture for volumetric segmentation, a challenging task that requires keeping a complex balance in encoding local and global spatial cues, and preserving information along all axes of the volume. Encoder of the proposed design benefits from self-attention mechanism to simultaneously encode local and global cues, while the decoder employs a parallel self and cross attention formulation to capture fine details for boundary refinement. Empirically, we show that the proposed design choices result in a computationally efficient model, with competitive and promising results on the Medical Segmentation Decathlon (MSD) brain tumor segmentation (BraTS) Task. We further show that the representations learned by our model are robust against data corruptions. Our code implementation is publicly available.

Original language	English
Title of host publication	Medical Image Computing and Computer Assisted Intervention – MICCAI 2022 - 25th International Conference Singapore, September 18–22, 2022 Proceedings, Part V
Editors	Linwei Wang, Qi Dou, P. Thomas Fletcher, Stefanie Speidel, Shuo Li
Place of Publication	Cham Switzerland
Publisher	Springer
Pages	162-172
Number of pages	11
ISBN (Electronic)	9783031164439
ISBN (Print)	9783031164422
DOIs	https://doi.org/10.1007/978-3-031-16443-9_16
Publication status	Published - 2022
Event	Medical Image Computing and Computer-Assisted Intervention 2022 - Singapore, Singapore Duration: 18 Sept 2022 → 22 Sept 2022 Conference number: 25th https://link.springer.com/book/10.1007/978-3-031-16434-7 (Proceedings - Part 2) https://conferences.miccai.org/2022/en/ (Website)

Publication series

Name	Lecture Notes in Computer Science
Publisher	Springer
Volume	13435
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	Medical Image Computing and Computer-Assisted Intervention 2022
Abbreviated title	MICCAI 2022
Country/Territory	Singapore
City	Singapore
Period	18/09/22 → 22/09/22
Internet address	https://link.springer.com/book/10.1007/978-3-031-16434-7 (Proceedings - Part 2) https://conferences.miccai.org/2022/en/ (Website)

Keywords

Pure volumetric transformer
Tumor segmentation

Access to Document

10.1007/978-3-031-16443-9_16

Cite this

Peiris, H., Hayat, M., Chen, Z., Egan, G., & Harandi, M. (2022). A robust volumetric Transformer for accurate 3D tumor segmentation. In L. Wang, Q. Dou, P. T. Fletcher, S. Speidel, & S. Li (Eds.), Medical Image Computing and Computer Assisted Intervention – MICCAI 2022 - 25th International Conference Singapore, September 18–22, 2022 Proceedings, Part V (pp. 162-172). (Lecture Notes in Computer Science; Vol. 13435). Springer. https://doi.org/10.1007/978-3-031-16443-9_16

Peiris, Himashi ; Hayat, Munawar ; Chen, Zhaolin et al. / A robust volumetric Transformer for accurate 3D tumor segmentation. Medical Image Computing and Computer Assisted Intervention – MICCAI 2022 - 25th International Conference Singapore, September 18–22, 2022 Proceedings, Part V. editor / Linwei Wang ; Qi Dou ; P. Thomas Fletcher ; Stefanie Speidel ; Shuo Li. Cham Switzerland : Springer, 2022. pp. 162-172 (Lecture Notes in Computer Science).

@inproceedings{d3cedc56213e4d8d85068c470a02b063,

title = "A robust volumetric Transformer for accurate 3D tumor segmentation",

abstract = "We propose a Transformer architecture for volumetric segmentation, a challenging task that requires keeping a complex balance in encoding local and global spatial cues, and preserving information along all axes of the volume. Encoder of the proposed design benefits from self-attention mechanism to simultaneously encode local and global cues, while the decoder employs a parallel self and cross attention formulation to capture fine details for boundary refinement. Empirically, we show that the proposed design choices result in a computationally efficient model, with competitive and promising results on the Medical Segmentation Decathlon (MSD) brain tumor segmentation (BraTS) Task. We further show that the representations learned by our model are robust against data corruptions. Our code implementation is publicly available.",

keywords = "Pure volumetric transformer, Tumor segmentation",

author = "Himashi Peiris and Munawar Hayat and Zhaolin Chen and Gary Egan and Mehrtash Harandi",

note = "Publisher Copyright: {\textcopyright} 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.; Medical Image Computing and Computer-Assisted Intervention 2022, MICCAI 2022 ; Conference date: 18-09-2022 Through 22-09-2022",

year = "2022",

doi = "10.1007/978-3-031-16443-9_16",

language = "English",

isbn = "9783031164422",

series = "Lecture Notes in Computer Science",

publisher = "Springer",

pages = "162--172",

editor = "Linwei Wang and Qi Dou and Fletcher, {P. Thomas} and Stefanie Speidel and Shuo Li",

booktitle = "Medical Image Computing and Computer Assisted Intervention – MICCAI 2022 - 25th International Conference Singapore, September 18–22, 2022 Proceedings, Part V",

url = "https://link.springer.com/book/10.1007/978-3-031-16434-7, https://conferences.miccai.org/2022/en/",

}

Peiris, H, Hayat, M, Chen, Z , Egan, G & Harandi, M 2022, A robust volumetric Transformer for accurate 3D tumor segmentation. in L Wang, Q Dou, PT Fletcher, S Speidel & S Li (eds), Medical Image Computing and Computer Assisted Intervention – MICCAI 2022 - 25th International Conference Singapore, September 18–22, 2022 Proceedings, Part V. Lecture Notes in Computer Science, vol. 13435, Springer, Cham Switzerland, pp. 162-172, Medical Image Computing and Computer-Assisted Intervention 2022, Singapore, Singapore, 18/09/22. https://doi.org/10.1007/978-3-031-16443-9_16

A robust volumetric Transformer for accurate 3D tumor segmentation. / Peiris, Himashi; Hayat, Munawar; Chen, Zhaolin et al.
Medical Image Computing and Computer Assisted Intervention – MICCAI 2022 - 25th International Conference Singapore, September 18–22, 2022 Proceedings, Part V. ed. / Linwei Wang; Qi Dou; P. Thomas Fletcher; Stefanie Speidel; Shuo Li. Cham Switzerland: Springer, 2022. p. 162-172 (Lecture Notes in Computer Science; Vol. 13435).

Research output: Chapter in Book/Report/Conference proceeding › Conference Paper › Research › peer-review

TY - GEN

T1 - A robust volumetric Transformer for accurate 3D tumor segmentation

AU - Peiris, Himashi

AU - Hayat, Munawar

AU - Chen, Zhaolin

AU - Egan, Gary

AU - Harandi, Mehrtash

N1 - Conference code: 25th

PY - 2022

Y1 - 2022

N2 - We propose a Transformer architecture for volumetric segmentation, a challenging task that requires keeping a complex balance in encoding local and global spatial cues, and preserving information along all axes of the volume. Encoder of the proposed design benefits from self-attention mechanism to simultaneously encode local and global cues, while the decoder employs a parallel self and cross attention formulation to capture fine details for boundary refinement. Empirically, we show that the proposed design choices result in a computationally efficient model, with competitive and promising results on the Medical Segmentation Decathlon (MSD) brain tumor segmentation (BraTS) Task. We further show that the representations learned by our model are robust against data corruptions. Our code implementation is publicly available.

AB - We propose a Transformer architecture for volumetric segmentation, a challenging task that requires keeping a complex balance in encoding local and global spatial cues, and preserving information along all axes of the volume. Encoder of the proposed design benefits from self-attention mechanism to simultaneously encode local and global cues, while the decoder employs a parallel self and cross attention formulation to capture fine details for boundary refinement. Empirically, we show that the proposed design choices result in a computationally efficient model, with competitive and promising results on the Medical Segmentation Decathlon (MSD) brain tumor segmentation (BraTS) Task. We further show that the representations learned by our model are robust against data corruptions. Our code implementation is publicly available.

KW - Pure volumetric transformer

KW - Tumor segmentation

UR - http://www.scopus.com/inward/record.url?scp=85139033880&partnerID=8YFLogxK

U2 - 10.1007/978-3-031-16443-9_16

DO - 10.1007/978-3-031-16443-9_16

M3 - Conference Paper

AN - SCOPUS:85139033880

SN - 9783031164422

T3 - Lecture Notes in Computer Science

SP - 162

EP - 172

BT - Medical Image Computing and Computer Assisted Intervention – MICCAI 2022 - 25th International Conference Singapore, September 18–22, 2022 Proceedings, Part V

A2 - Wang, Linwei

A2 - Dou, Qi

A2 - Fletcher, P. Thomas

A2 - Speidel, Stefanie

A2 - Li, Shuo

PB - Springer

CY - Cham Switzerland

T2 - Medical Image Computing and Computer-Assisted Intervention 2022

Y2 - 18 September 2022 through 22 September 2022

ER -

Peiris H, Hayat M, Chen Z , Egan G , Harandi M. A robust volumetric Transformer for accurate 3D tumor segmentation. In Wang L, Dou Q, Fletcher PT, Speidel S, Li S, editors, Medical Image Computing and Computer Assisted Intervention – MICCAI 2022 - 25th International Conference Singapore, September 18–22, 2022 Proceedings, Part V. Cham Switzerland: Springer. 2022. p. 162-172. (Lecture Notes in Computer Science). doi: 10.1007/978-3-031-16443-9_16