MixLacune: Segmentation of lacunes of presumed vascular origin

Denis Kutnar; Bas van der Velden; Marta Girones Sanguesa; Mirjam I Geerlings; Matthijs Biesbroek; Hugo Kuijf

doi:10.48550/arXiv.2108.02483

MixLacune: Segmentation of lacunes of presumed vascular origin

Denis Kutnar, Bas van der Velden, Marta Girones Sanguesa, Mirjam I Geerlings, Matthijs Biesbroek, Hugo Kuijf

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic

10 Downloads (Pure)

Abstract

Lacunes of presumed vascular origin are fluid-filled cavities of between 3 - 15 mm in diameter, visible on T1 and FLAIR brain MRI. Quantification of lacunes relies on manual annotation or semi-automatic / interactive approaches; and almost no automatic methods exist for this task. In this work, we present a two-stage approach to segment lacunes of presumed vascular origin: (1) detection with Mask R-CNN followed by (2) segmentation with a U-Net CNN. Data originates from Task 3 of the "Where is VALDO?" challenge and consists of 40 training subjects. We report the mean DICE on the training set of 0.83 and on the validation set of 0.84. Source code is available at: https://github.com/hjkuijf/MixLacune . The docker container hjkuijf/mixlacune can be pulled from https://hub.docker.com/r/hjkuijf/mixlacune .

Original language	English
Title of host publication	"Where is VALDO?" challenge, MICCAI 2021
Number of pages	10
DOIs	https://doi.org/10.48550/arXiv.2108.02483
Publication status	Published - 2021

Access to Document

10.48550/arXiv.2108.02483Licence: CC BY

2108.02483Final published version, 8.3 MBLicence: CC BY

Cite this

@inproceedings{64f5fd211df34d309e66daae5bd81aae,

title = "MixLacune: Segmentation of lacunes of presumed vascular origin",

abstract = "Lacunes of presumed vascular origin are fluid-filled cavities of between 3 - 15 mm in diameter, visible on T1 and FLAIR brain MRI. Quantification of lacunes relies on manual annotation or semi-automatic / interactive approaches; and almost no automatic methods exist for this task. In this work, we present a two-stage approach to segment lacunes of presumed vascular origin: (1) detection with Mask R-CNN followed by (2) segmentation with a U-Net CNN. Data originates from Task 3 of the {"}Where is VALDO?{"} challenge and consists of 40 training subjects. We report the mean DICE on the training set of 0.83 and on the validation set of 0.84. Source code is available at: https://github.com/hjkuijf/MixLacune . The docker container hjkuijf/mixlacune can be pulled from https://hub.docker.com/r/hjkuijf/mixlacune .",

author = "Denis Kutnar and {van der Velden}, Bas and {Girones Sanguesa}, Marta and Geerlings, {Mirjam I} and Matthijs Biesbroek and Hugo Kuijf",

year = "2021",

doi = "10.48550/arXiv.2108.02483",

language = "English",

booktitle = "{"}Where is VALDO?{"} challenge, MICCAI 2021",

}

TY - GEN

T1 - MixLacune: Segmentation of lacunes of presumed vascular origin

AU - Kutnar, Denis

AU - van der Velden, Bas

AU - Girones Sanguesa, Marta

AU - Geerlings, Mirjam I

AU - Biesbroek, Matthijs

AU - Kuijf, Hugo

PY - 2021

Y1 - 2021

N2 - Lacunes of presumed vascular origin are fluid-filled cavities of between 3 - 15 mm in diameter, visible on T1 and FLAIR brain MRI. Quantification of lacunes relies on manual annotation or semi-automatic / interactive approaches; and almost no automatic methods exist for this task. In this work, we present a two-stage approach to segment lacunes of presumed vascular origin: (1) detection with Mask R-CNN followed by (2) segmentation with a U-Net CNN. Data originates from Task 3 of the "Where is VALDO?" challenge and consists of 40 training subjects. We report the mean DICE on the training set of 0.83 and on the validation set of 0.84. Source code is available at: https://github.com/hjkuijf/MixLacune . The docker container hjkuijf/mixlacune can be pulled from https://hub.docker.com/r/hjkuijf/mixlacune .

AB - Lacunes of presumed vascular origin are fluid-filled cavities of between 3 - 15 mm in diameter, visible on T1 and FLAIR brain MRI. Quantification of lacunes relies on manual annotation or semi-automatic / interactive approaches; and almost no automatic methods exist for this task. In this work, we present a two-stage approach to segment lacunes of presumed vascular origin: (1) detection with Mask R-CNN followed by (2) segmentation with a U-Net CNN. Data originates from Task 3 of the "Where is VALDO?" challenge and consists of 40 training subjects. We report the mean DICE on the training set of 0.83 and on the validation set of 0.84. Source code is available at: https://github.com/hjkuijf/MixLacune . The docker container hjkuijf/mixlacune can be pulled from https://hub.docker.com/r/hjkuijf/mixlacune .

U2 - 10.48550/arXiv.2108.02483

DO - 10.48550/arXiv.2108.02483

M3 - Conference contribution

BT - "Where is VALDO?" challenge, MICCAI 2021

ER -

MixLacune: Segmentation of lacunes of presumed vascular origin

Abstract

Access to Document

Fingerprint

Cite this