Weakly Supervised Volumetric Image Segmentation with Deformed Templates

Udaranga Wickramasinghe*, Patrick Jensen, Mian Shah, Jiancheng Yang, Pascal Fua

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

29 Downloads (Pure)

Abstract

There are many approaches to weakly-supervised training of networks to segment 2D images. By contrast, existing approaches to segmenting volumetric images rely on full-supervision of a subset of 2D slices of the 3D volume. We propose an approach to volume segmentation that is truly weakly-supervised in the sense that we only need to provide a sparse set of 3D points on the surface of target objects instead of detailed 2D masks. We use the 3D points to deform a 3D template so that it roughly matches the target object outlines and we introduce an architecture that exploits the supervision it provides to train a network to find accurate boundaries. We evaluate our approach on Computed Tomography (CT), Magnetic Resonance Imagery (MRI) and Electron Microscopy (EM) image datasets and show that it substantially reduces the required amount of effort.

Original languageEnglish
Title of host publicationProceedings of the 25th International Conference of Medical Image Computing and Computer Assisted Intervention, MICCAI 2022
Volume13435
PublisherSpringer
Publication date2022
Pages422-432
ISBN (Print) 978-3-031-16442-2
ISBN (Electronic)978-3-031-16443-9
DOIs
Publication statusPublished - 2022
Event25th International Conference on Medical Image Computing and Computer-Assisted Intervention - Resorts World Convention Centre Singapore, Singapore, Singapore
Duration: 18 Sept 202222 Sept 2022
Conference number: 25
https://conferences.miccai.org/2022/en/

Conference

Conference25th International Conference on Medical Image Computing and Computer-Assisted Intervention
Number25
LocationResorts World Convention Centre Singapore
Country/TerritorySingapore
CitySingapore
Period18/09/202222/09/2022
Internet address
SeriesLecture Notes in Computer Science
Volume13435
ISSN0302-9743

Fingerprint

Dive into the research topics of 'Weakly Supervised Volumetric Image Segmentation with Deformed Templates'. Together they form a unique fingerprint.

Cite this