Is Segmentation Uncertainty Useful?

Steffen Czolbe, Kasra Arnavaz, Oswin Krause, Aasa Feragen

Research output: Chapter in Book/Report/Conference proceedingBook chapterResearchpeer-review

Abstract

Probabilistic image segmentation encodes varying prediction confidence and inherent ambiguity in the segmentation problem. While different probabilistic segmentation models are designed to capture different aspects of segmentation uncertainty and ambiguity, these modelling differences are rarely discussed in the context of applications of uncertainty. We consider two common use cases of segmentation uncertainty, namely assessment of segmentation quality and active learning. We consider four established strategies for probabilistic segmentation, discuss their modelling capabilities, and investigate their performance in these two tasks. We find that for all models and both tasks, returned uncertainty correlates positively with segmentation error, but does not prove to be useful for active learning.
Original languageEnglish
Title of host publicationInformation Processing in Medical Imaging
PublisherSpringer
Publication date2021
Pages715-726
ISBN (Print)978-3-030-78190-3
DOIs
Publication statusPublished - 2021
EventInternational Conference on Information Processing in Medical Imaging - Virtual Event
Duration: 28 Jun 202130 Jun 2021

Conference

ConferenceInternational Conference on Information Processing in Medical Imaging
LocationVirtual Event
Period28/06/202130/06/2021
SeriesLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume12729
ISSN0302-9743

Keywords

  • Active learning
  • Image segmentation
  • Uncertainty quantification

Fingerprint

Dive into the research topics of 'Is Segmentation Uncertainty Useful?'. Together they form a unique fingerprint.

Cite this