Improving Aleatoric Uncertainty Quantification in Multi-annotated Medical Image Segmentation with Normalizing Flows

Onderzoeksoutput: Hoofdstuk in Boek/Rapport/CongresprocedureConferentiebijdrageAcademicpeer review

12 Citaten (Scopus)
2 Downloads (Pure)

Samenvatting

Quantifying uncertainty in medical image segmentation applications is essential, as it is often connected to vital decision-making. Compelling attempts have been made in quantifying the uncertainty in image segmentation architectures, e.g. to learn a density segmentation model conditioned on the input image. Typical work in this field restricts these learnt densities to be strictly Gaussian. In this paper, we propose to use a more flexible approach by introducing Normalizing Flows (NFs), which enables the learnt densities to be more complex and facilitate more accurate modeling for uncertainty. We prove this hypothesis by adopting the Probabilistic U-Net and augmenting the posterior density with an NF, allowing it to be more expressive. Our qualitative as well as quantitative (GED and IoU) evaluations on the multi-annotated and single-annotated LIDC-IDRI and Kvasir-SEG segmentation datasets, respectively, show a clear improvement. This is mostly apparent in the quantification of aleatoric uncertainty and the increased predictive performance of up to 14%. This result strongly indicates that a more flexible density model should be seriously considered in architectures that attempt to capture segmentation ambiguity through density modeling. The benefit of this improved modeling will increase human confidence in annotation and segmentation, and enable eager adoption of the technology in practice.

Originele taal-2Engels
TitelUncertainty for Safe Utilization of Machine Learning in Medical Imaging, and Perinatal Imaging, Placental and Preterm Image Analysis
Subtitel3rd International Workshop, UNSURE 2021, and 6th International Workshop, PIPPI 2021, Held in Conjunction with MICCAI 2021, Strasbourg, France, October 1, 2021, Proceedings
RedacteurenCarole H. Sudre, Roxane Licandro, Christian Baumgartner, Andrew Melbourne, Adrian Dalca, Jana Hutter, Ryutaro Tanno, Esra Abaci Turk, Koen Van Leemput, Jordina Torrents Barrena, William M. Wells, Christopher Macgowan
Plaats van productieCham
UitgeverijSpringer
Hoofdstuk8
Pagina's75-88
Aantal pagina's14
ISBN van elektronische versie978-3-030-87735-4
ISBN van geprinte versie978-3-030-87734-7
DOI's
StatusGepubliceerd - 2021
Evenement3rd International Workshop on Uncertainty for Safe Utilization of Machine Learning in Medical Imaging, UNSURE 2021, and the 6th International Workshop on Perinatal, Preterm and Paediatric Image Analysis, PIPPI 2021, held in conjunction with the 24th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2021 - Virtual, Online, Strasbourg, Frankrijk
Duur: 1 okt. 20211 okt. 2021

Publicatie series

NaamLecture Notes in Computer Science (LNCS)
UitgeverijSpringer
Volume12959
ISSN van geprinte versie0302-9743
ISSN van elektronische versie1611-3349
NaamImage Processing, Computer Vision, Pattern Recognition, and Graphics (LNIP)
UitgeverijSpringer
Volume12959

Congres

Congres3rd International Workshop on Uncertainty for Safe Utilization of Machine Learning in Medical Imaging, UNSURE 2021, and the 6th International Workshop on Perinatal, Preterm and Paediatric Image Analysis, PIPPI 2021, held in conjunction with the 24th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2021
Land/RegioFrankrijk
StadStrasbourg
Periode1/10/211/10/21

Bibliografische nota

Publisher Copyright:
© 2021, Springer Nature Switzerland AG.

Vingerafdruk

Duik in de onderzoeksthema's van 'Improving Aleatoric Uncertainty Quantification in Multi-annotated Medical Image Segmentation with Normalizing Flows'. Samen vormen ze een unieke vingerafdruk.

Citeer dit