N3 bias field correction explained as a Bayesian modeling method

Christian Thode Larsen*, J. Eugenio Iglesias, Koen Van Leemput

*Corresponding author for this work

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

    18 Citations (Scopus)

    Abstract

    Although N3 is perhaps the most widely used method for MRI bias field correction, its underlying mechanism is in fact not well understood. Specifically, the method relies on a relatively heuristic recipe of alternating iterative steps that does not optimize any particular objective function. In this paper we explain the successful bias field correction properties of N3 by showing that it implicitly uses the same generative models and computational strategies as expectation maximization (EM) based bias field correction methods.We demonstrate experimentally that purely EM-based methods are capable of producing bias field correction results comparable to those of N3 in less computation time.

    Original languageEnglish
    Title of host publicationBayesian and grAphical Models for Biomedical Imaging
    Subtitle of host publicationFirst International Workshop, BAMBI 2014, Cambridge, MA, USA, September 18, 2014, revised selected papers
    EditorsM. Jorge Cardoso, Ivor Simpson, Tal Arbel, Doina Precup, Annemie Ribbens
    PublisherSpringer
    Pages1-12
    Number of pages12
    Volume8677
    ISBN (Electronic)978-3-319-12289-2
    ISBN (Print)978-3-319-12288-5
    DOIs
    Publication statusPublished - 2014
    MoE publication typeA4 Conference publication
    EventInternational Workshop on Bayesian and grAphical Models for Biomedical Imaging - Cambridge, United States
    Duration: 18 Sept 201418 Sept 2014
    Conference number: 1

    Publication series

    NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
    PublisherSpringer
    Volume8677
    ISSN (Print)0302-9743
    ISSN (Electronic)1611-3349

    Workshop

    WorkshopInternational Workshop on Bayesian and grAphical Models for Biomedical Imaging
    Abbreviated titleBAMBI
    Country/TerritoryUnited States
    CityCambridge
    Period18/09/201418/09/2014

    Fingerprint

    Dive into the research topics of 'N3 bias field correction explained as a Bayesian modeling method'. Together they form a unique fingerprint.

    Cite this