Higher-order graph cuts and medical image segmentation

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Abstract

    Energy minimization is regularly used for medical image segmentation. Higher-order energies are perhaps not as common, but are nevertheless being used increasingly often. Whereas the common first- order (pairwise) potential can directly model only the relationship between pairs of pixels, the higher-order potential can model more complex and useful relationships between more than two variables. For instance, sets of pixels, chosen according to the shape to be segmented, can be encouraged to be entirely in one segment or the other by higher-order terms. In this talk, I will describe methods for minimizing higher-order energies using graph cuts as well as some real-world examples of their applications in medical image segmentation that have been deployed in commercial medical imaging software.

    Original languageEnglish
    Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
    PublisherSpringer Verlag
    Volume9555
    ISBN (Print)9783319302843
    Publication statusPublished - 2016
    Event7th Pacific-Rim Symposium on Image and Video Technology, PSIVT 2015 - Auckland, New Zealand
    Duration: 2015 Nov 232015 Nov 27

    Publication series

    NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
    Volume9555
    ISSN (Print)03029743
    ISSN (Electronic)16113349

    Other

    Other7th Pacific-Rim Symposium on Image and Video Technology, PSIVT 2015
    CountryNew Zealand
    CityAuckland
    Period15/11/2315/11/27

    Fingerprint

    Graph Cuts
    Medical Image
    Image segmentation
    Image Segmentation
    Pixels
    Higher Order
    Medical imaging
    Pixel
    Energy Minimization
    Medical Imaging
    Energy
    Pairwise
    First-order
    Software
    Term
    Model
    Relationships

    Keywords

    • Graph cuts
    • Graphical model
    • Higher-order energy
    • Markov random fields
    • MRF
    • Segmentation

    ASJC Scopus subject areas

    • Computer Science(all)
    • Theoretical Computer Science

    Cite this

    Ishikawa, H. (2016). Higher-order graph cuts and medical image segmentation. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 9555). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9555). Springer Verlag.

    Higher-order graph cuts and medical image segmentation. / Ishikawa, Hiroshi.

    Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 9555 Springer Verlag, 2016. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9555).

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Ishikawa, H 2016, Higher-order graph cuts and medical image segmentation. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 9555, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9555, Springer Verlag, 7th Pacific-Rim Symposium on Image and Video Technology, PSIVT 2015, Auckland, New Zealand, 15/11/23.
    Ishikawa H. Higher-order graph cuts and medical image segmentation. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 9555. Springer Verlag. 2016. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
    Ishikawa, Hiroshi. / Higher-order graph cuts and medical image segmentation. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 9555 Springer Verlag, 2016. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
    @inproceedings{d5817496682d4169afc7a32e337a403b,
    title = "Higher-order graph cuts and medical image segmentation",
    abstract = "Energy minimization is regularly used for medical image segmentation. Higher-order energies are perhaps not as common, but are nevertheless being used increasingly often. Whereas the common first- order (pairwise) potential can directly model only the relationship between pairs of pixels, the higher-order potential can model more complex and useful relationships between more than two variables. For instance, sets of pixels, chosen according to the shape to be segmented, can be encouraged to be entirely in one segment or the other by higher-order terms. In this talk, I will describe methods for minimizing higher-order energies using graph cuts as well as some real-world examples of their applications in medical image segmentation that have been deployed in commercial medical imaging software.",
    keywords = "Graph cuts, Graphical model, Higher-order energy, Markov random fields, MRF, Segmentation",
    author = "Hiroshi Ishikawa",
    year = "2016",
    language = "English",
    isbn = "9783319302843",
    volume = "9555",
    series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
    publisher = "Springer Verlag",
    booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

    }

    TY - GEN

    T1 - Higher-order graph cuts and medical image segmentation

    AU - Ishikawa, Hiroshi

    PY - 2016

    Y1 - 2016

    N2 - Energy minimization is regularly used for medical image segmentation. Higher-order energies are perhaps not as common, but are nevertheless being used increasingly often. Whereas the common first- order (pairwise) potential can directly model only the relationship between pairs of pixels, the higher-order potential can model more complex and useful relationships between more than two variables. For instance, sets of pixels, chosen according to the shape to be segmented, can be encouraged to be entirely in one segment or the other by higher-order terms. In this talk, I will describe methods for minimizing higher-order energies using graph cuts as well as some real-world examples of their applications in medical image segmentation that have been deployed in commercial medical imaging software.

    AB - Energy minimization is regularly used for medical image segmentation. Higher-order energies are perhaps not as common, but are nevertheless being used increasingly often. Whereas the common first- order (pairwise) potential can directly model only the relationship between pairs of pixels, the higher-order potential can model more complex and useful relationships between more than two variables. For instance, sets of pixels, chosen according to the shape to be segmented, can be encouraged to be entirely in one segment or the other by higher-order terms. In this talk, I will describe methods for minimizing higher-order energies using graph cuts as well as some real-world examples of their applications in medical image segmentation that have been deployed in commercial medical imaging software.

    KW - Graph cuts

    KW - Graphical model

    KW - Higher-order energy

    KW - Markov random fields

    KW - MRF

    KW - Segmentation

    UR - http://www.scopus.com/inward/record.url?scp=84960443966&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=84960443966&partnerID=8YFLogxK

    M3 - Conference contribution

    SN - 9783319302843

    VL - 9555

    T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

    BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

    PB - Springer Verlag

    ER -