3D CMM-Net with Deeper Encoder for Semantic Segmentation of Brain Tumors in BraTS2021 Challenge

Yoonseok Choi, Mohammed A. Al-masni, Dong Hyun Kim

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

We propose a 3D version of the Contextual Multi-scale Multi-level Network (3D CMM-Net) with deeper encoder depth for automated semantic segmentation of different brain tumors in the BraTS2021 challenge. The proposed network has the capability to extract and learn deeper features for the task of multi-class segmentation directly from 3D MRI data. The overall performance of the proposed network gave Dice scores of 0.7557, 0.8060, and 0.8351 for enhancing tumor, tumor core, and whole tumor, respectively on the local-test dataset.

Original languageEnglish
Title of host publicationBrainlesion
Subtitle of host publicationGlioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries - 7th International Workshop, BrainLes 2021, Held in Conjunction with MICCAI 2021, Revised Selected Papers
EditorsAlessandro Crimi, Spyridon Bakas
PublisherSpringer Science and Business Media Deutschland GmbH
Pages333-343
Number of pages11
ISBN (Print)9783031089985
DOIs
StatePublished - 2022
Event7th International Brain Lesion Workshop, BrainLes 2021, held in conjunction with the Medical Image Computing and Computer Assisted Intervention, MICCAI 2021 - Virtual, Online
Duration: 27 Sep 202127 Sep 2021

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume12962 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference7th International Brain Lesion Workshop, BrainLes 2021, held in conjunction with the Medical Image Computing and Computer Assisted Intervention, MICCAI 2021
CityVirtual, Online
Period27/09/2127/09/21

Keywords

  • 3D semantic segmentation
  • Brain tumor segmentation
  • Glioblastoma
  • Multimodal MRI
  • Pyramid pooling module
  • U-Net

Fingerprint

Dive into the research topics of '3D CMM-Net with Deeper Encoder for Semantic Segmentation of Brain Tumors in BraTS2021 Challenge'. Together they form a unique fingerprint.

Cite this