Leveraging Ellipsoid Bounding Shapes and Fast R-CNN for Enlarged Perivascular Spaces Detection and Segmentation

Mariam Zabihi; Chayanin Tangwiriyasakul; Silvia Ingala; Luigi Lorenzini; Robin Camarasa; Frederik Barkhof; Marleen de Bruijne; M. Jorge Cardoso; Carole H. Sudre

doi:10.1007/978-3-031-45676-3_33

Leveraging Ellipsoid Bounding Shapes and Fast R-CNN for Enlarged Perivascular Spaces Detection and Segmentation

Mariam Zabihi^*, Chayanin Tangwiriyasakul, Silvia Ingala, Luigi Lorenzini, Robin Camarasa, Frederik Barkhof, Marleen de Bruijne, M. Jorge Cardoso, Carole H. Sudre

^*Corresponding author for this work

Department of Diagnostic Radiology

Abstract

Enlarged perivascular spaces (EPVS) are small fluid-filled spaces surrounding blood vessels in the brain. They have been found to be important in the development and progression of cerebrovascular disease, including stroke, dementia, and cerebral small vessel disease. Their accurate detection and quantification are crucial for early diagnosis and better management of these diseases. In recent years, object detection techniques such as Mask R-CNN approach have been widely used to automate the detection and segmentation of small objects. To account for the tubular shape of these markers we use ellipsoid shapes instead of bounding boxes to express the location of individual elements in the implementation of the Fast R-CNN. We investigate the performance of this model under different modality combinations and find that the T2 modality alone, as well as the combination of T1+T2, deliver better performance.

Original language	English
Title of host publication	Machine Learning in Medical Imaging - 14th International Workshop, MLMI 2023, Held in Conjunction with MICCAI 2023, Proceedings
Editors	Xiaohuan Cao, Xi Ouyang, Xuanang Xu, Islem Rekik, Zhiming Cui
Number of pages	10
Publisher	Springer Science and Business Media Deutschland GmbH
Publication date	2024
Pages	325-334
ISBN (Print)	9783031456756
DOIs	https://doi.org/10.1007/978-3-031-45676-3_33
Publication status	Published - 2024
Event	14th International Workshop on Machine Learning in Medical Imaging, MLMI 2023 - Vancouver, Canada Duration: 8 Oct 2023 → 8 Oct 2023

Conference

Conference	14th International Workshop on Machine Learning in Medical Imaging, MLMI 2023
Country/Territory	Canada
City	Vancouver
Period	08/10/2023 → 08/10/2023

Series	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	14349 LNCS
ISSN	0302-9743

Keywords

Cerebrovascular diseases
Ellipsoid bounding shapes
enlarged perivascular spaces
Fast R-CNN

Access to Document

10.1007/978-3-031-45676-3_33

Cite this

Zabihi, M., Tangwiriyasakul, C., Ingala, S., Lorenzini, L., Camarasa, R., Barkhof, F., de Bruijne, M., Cardoso, M. J., & Sudre, C. H. (2024). Leveraging Ellipsoid Bounding Shapes and Fast R-CNN for Enlarged Perivascular Spaces Detection and Segmentation. In X. Cao, X. Ouyang, X. Xu, I. Rekik, & Z. Cui (Eds.), Machine Learning in Medical Imaging - 14th International Workshop, MLMI 2023, Held in Conjunction with MICCAI 2023, Proceedings (pp. 325-334). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-031-45676-3_33

Leveraging Ellipsoid Bounding Shapes and Fast R-CNN for Enlarged Perivascular Spaces Detection and Segmentation. / Zabihi, Mariam; Tangwiriyasakul, Chayanin; Ingala, Silvia et al.
Machine Learning in Medical Imaging - 14th International Workshop, MLMI 2023, Held in Conjunction with MICCAI 2023, Proceedings. ed. / Xiaohuan Cao; Xi Ouyang; Xuanang Xu; Islem Rekik; Zhiming Cui. Springer Science and Business Media Deutschland GmbH, 2024. p. 325-334 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Vol. 14349 LNCS).

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

Zabihi, M, Tangwiriyasakul, C, Ingala, S, Lorenzini, L, Camarasa, R, Barkhof, F, de Bruijne, M, Cardoso, MJ & Sudre, CH 2024, Leveraging Ellipsoid Bounding Shapes and Fast R-CNN for Enlarged Perivascular Spaces Detection and Segmentation. in X Cao, X Ouyang, X Xu, I Rekik & Z Cui (eds), Machine Learning in Medical Imaging - 14th International Workshop, MLMI 2023, Held in Conjunction with MICCAI 2023, Proceedings. Springer Science and Business Media Deutschland GmbH, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 14349 LNCS, pp. 325-334, 14th International Workshop on Machine Learning in Medical Imaging, MLMI 2023, Vancouver, Canada, 08/10/2023. https://doi.org/10.1007/978-3-031-45676-3_33

Zabihi M, Tangwiriyasakul C, Ingala S, Lorenzini L, Camarasa R, Barkhof F et al. Leveraging Ellipsoid Bounding Shapes and Fast R-CNN for Enlarged Perivascular Spaces Detection and Segmentation. In Cao X, Ouyang X, Xu X, Rekik I, Cui Z, editors, Machine Learning in Medical Imaging - 14th International Workshop, MLMI 2023, Held in Conjunction with MICCAI 2023, Proceedings. Springer Science and Business Media Deutschland GmbH. 2024. p. 325-334. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Vol. 14349 LNCS). doi: 10.1007/978-3-031-45676-3_33

Zabihi, Mariam ; Tangwiriyasakul, Chayanin ; Ingala, Silvia et al. / Leveraging Ellipsoid Bounding Shapes and Fast R-CNN for Enlarged Perivascular Spaces Detection and Segmentation. Machine Learning in Medical Imaging - 14th International Workshop, MLMI 2023, Held in Conjunction with MICCAI 2023, Proceedings. editor / Xiaohuan Cao ; Xi Ouyang ; Xuanang Xu ; Islem Rekik ; Zhiming Cui. Springer Science and Business Media Deutschland GmbH, 2024. pp. 325-334 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Vol. 14349 LNCS).

@inproceedings{7c4e1ee3f5854b688985e0be0ec57d25,

title = "Leveraging Ellipsoid Bounding Shapes and Fast R-CNN for Enlarged Perivascular Spaces Detection and Segmentation",

abstract = "Enlarged perivascular spaces (EPVS) are small fluid-filled spaces surrounding blood vessels in the brain. They have been found to be important in the development and progression of cerebrovascular disease, including stroke, dementia, and cerebral small vessel disease. Their accurate detection and quantification are crucial for early diagnosis and better management of these diseases. In recent years, object detection techniques such as Mask R-CNN approach have been widely used to automate the detection and segmentation of small objects. To account for the tubular shape of these markers we use ellipsoid shapes instead of bounding boxes to express the location of individual elements in the implementation of the Fast R-CNN. We investigate the performance of this model under different modality combinations and find that the T2 modality alone, as well as the combination of T1+T2, deliver better performance.",

keywords = "Cerebrovascular diseases, Ellipsoid bounding shapes, enlarged perivascular spaces, Fast R-CNN",

author = "Mariam Zabihi and Chayanin Tangwiriyasakul and Silvia Ingala and Luigi Lorenzini and Robin Camarasa and Frederik Barkhof and {de Bruijne}, Marleen and Cardoso, {M. Jorge} and Sudre, {Carole H.}",

note = "Publisher Copyright: {\textcopyright} 2024, The Author(s), under exclusive license to Springer Nature Switzerland AG.; 14th International Workshop on Machine Learning in Medical Imaging, MLMI 2023 ; Conference date: 08-10-2023 Through 08-10-2023",

year = "2024",

doi = "10.1007/978-3-031-45676-3_33",

language = "English",

isbn = "9783031456756",

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

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "325--334",

editor = "Xiaohuan Cao and Xi Ouyang and Xuanang Xu and Islem Rekik and Zhiming Cui",

booktitle = "Machine Learning in Medical Imaging - 14th International Workshop, MLMI 2023, Held in Conjunction with MICCAI 2023, Proceedings",

address = "Germany",

}

TY - GEN

T1 - Leveraging Ellipsoid Bounding Shapes and Fast R-CNN for Enlarged Perivascular Spaces Detection and Segmentation

AU - Zabihi, Mariam

AU - Tangwiriyasakul, Chayanin

AU - Ingala, Silvia

AU - Lorenzini, Luigi

AU - Camarasa, Robin

AU - Barkhof, Frederik

AU - de Bruijne, Marleen

AU - Cardoso, M. Jorge

AU - Sudre, Carole H.

PY - 2024

Y1 - 2024

N2 - Enlarged perivascular spaces (EPVS) are small fluid-filled spaces surrounding blood vessels in the brain. They have been found to be important in the development and progression of cerebrovascular disease, including stroke, dementia, and cerebral small vessel disease. Their accurate detection and quantification are crucial for early diagnosis and better management of these diseases. In recent years, object detection techniques such as Mask R-CNN approach have been widely used to automate the detection and segmentation of small objects. To account for the tubular shape of these markers we use ellipsoid shapes instead of bounding boxes to express the location of individual elements in the implementation of the Fast R-CNN. We investigate the performance of this model under different modality combinations and find that the T2 modality alone, as well as the combination of T1+T2, deliver better performance.

AB - Enlarged perivascular spaces (EPVS) are small fluid-filled spaces surrounding blood vessels in the brain. They have been found to be important in the development and progression of cerebrovascular disease, including stroke, dementia, and cerebral small vessel disease. Their accurate detection and quantification are crucial for early diagnosis and better management of these diseases. In recent years, object detection techniques such as Mask R-CNN approach have been widely used to automate the detection and segmentation of small objects. To account for the tubular shape of these markers we use ellipsoid shapes instead of bounding boxes to express the location of individual elements in the implementation of the Fast R-CNN. We investigate the performance of this model under different modality combinations and find that the T2 modality alone, as well as the combination of T1+T2, deliver better performance.

KW - Cerebrovascular diseases

KW - Ellipsoid bounding shapes

KW - enlarged perivascular spaces

KW - Fast R-CNN

UR - http://www.scopus.com/inward/record.url?scp=85175945790&partnerID=8YFLogxK

U2 - 10.1007/978-3-031-45676-3_33

DO - 10.1007/978-3-031-45676-3_33

M3 - Article in proceedings

AN - SCOPUS:85175945790

SN - 9783031456756

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

SP - 325

EP - 334

BT - Machine Learning in Medical Imaging - 14th International Workshop, MLMI 2023, Held in Conjunction with MICCAI 2023, Proceedings

A2 - Cao, Xiaohuan

A2 - Ouyang, Xi

A2 - Xu, Xuanang

A2 - Rekik, Islem

A2 - Cui, Zhiming

PB - Springer Science and Business Media Deutschland GmbH

T2 - 14th International Workshop on Machine Learning in Medical Imaging, MLMI 2023

Y2 - 8 October 2023 through 8 October 2023

ER -