Deep Label Fusion: A 3D End-To-End Hybrid Multi-atlas Segmentation and Deep Learning Pipeline

Long Xie, Laura E.M. Wisse, Jiancong Wang, Sadhana Ravikumar, Trevor Glenn, Anica Luther, Sydney Lim, David A. Wolk, Paul A. Yushkevich

Forskningsoutput: Kapitel i bok/rapport/Conference proceedingKonferenspaper i proceedingPeer review

Sammanfattning

Deep learning (DL) is the state-of-the-art methodology in various medical image segmentation tasks. However, it requires relatively large amounts of manually labeled training data, which may be infeasible to generate in some applications. In addition, DL methods have relatively poor generalizability to out-of-sample data. Multi-atlas segmentation (MAS), on the other hand, has promising performance using limited amounts of training data and good generalizability. A hybrid method that integrates the high accuracy of DL and good generalizability of MAS is highly desired and could play an important role in segmentation problems where manually labeled data is hard to generate. Most of the prior work focuses on improving single components of MAS using DL rather than directly optimizing the final segmentation accuracy via an end-to-end pipeline. Only one study explored this idea in binary segmentation of 2D images, but it remains unknown whether it generalizes well to multi-class 3D segmentation problems. In this study, we propose a 3D end-to-end hybrid pipeline, named deep label fusion (DLF), that takes advantage of the strengths of MAS and DL. Experimental results demonstrate that DLF yields significant improvements over conventional label fusion methods and U-Net, a direct DL approach, in the context of segmenting medial temporal lobe subregions using 3T T1-weighted and T2-weighted MRI. Further, when applied to an unseen similar dataset acquired in 7T, DLF maintains its superior performance, which demonstrates its good generalizability.

Originalspråkengelska
Titel på gästpublikationInformation Processing in Medical Imaging - 27th International Conference, IPMI 2021, Proceedings
RedaktörerAasa Feragen, Stefan Sommer, Julia Schnabel, Mads Nielsen
FörlagSpringer Science and Business Media B.V.
Sidor428-439
Antal sidor12
ISBN (tryckt)9783030781903
DOI
StatusPublished - 2021
Evenemang27th International Conference on Information Processing in Medical Imaging, IPMI 2021 - Virtual, Online
Varaktighet: 2021 jun 282021 jun 30

Publikationsserier

NamnLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volym12729 LNCS
ISSN (tryckt)0302-9743
ISSN (elektroniskt)1611-3349

Konferens

Konferens27th International Conference on Information Processing in Medical Imaging, IPMI 2021
OrtVirtual, Online
Period2021/06/282021/06/30

Ämnesklassifikation (UKÄ)

  • Medicinsk bildbehandling

Fingeravtryck

Utforska forskningsämnen för ”Deep Label Fusion: A 3D End-To-End Hybrid Multi-atlas Segmentation and Deep Learning Pipeline”. Tillsammans bildar de ett unikt fingeravtryck.

Citera det här