Optimal data-driven sparse parameterization of diffeomorphisms for population analysis

Sandy Durrleman, Marcel Prastawa, Guido Gerig, Sarang Joshi

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

    Abstract

    In this paper, we propose a novel approach for intensity based atlas construction from a population of anatomical images, that estimates not only a template representative image but also a common optimal parameterization of the anatomical variations evident in the population. First, we introduce a discrete parameterization of large diffeomorphic deformations based on a finite set of control points, so that deformations are characterized by a low dimensional geometric descriptor. Second, we optimally estimate the position of the control points in the template image domain. As a consequence, control points move to where they are needed most to capture the geometric variability evident in the population. Third, the optimal number of control points is estimated by using a log - L1 sparsity penalty. The estimation of the template image, the template-to-subject mappings and their optimal parameterization is done via a single gradient descent optimization, and at the same computational cost as independent template-to-subject registrations. We present results that show that the anatomical variability of the population can be encoded efficiently with these compact and adapted geometric descriptors.

    Original languageEnglish (US)
    Title of host publicationInformation Processing in Medical Imaging - 22nd International Conference, IPMI 2011, Proceedings
    Pages123-134
    Number of pages12
    Volume6801 LNCS
    DOIs
    StatePublished - 2011
    Event22nd International Conference on Information Processing in Medical Imaging, IPMI 2011 - Kloster Irsee, Germany
    Duration: Jul 3 2011Jul 8 2011

    Publication series

    NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
    Volume6801 LNCS
    ISSN (Print)03029743
    ISSN (Electronic)16113349

    Other

    Other22nd International Conference on Information Processing in Medical Imaging, IPMI 2011
    CountryGermany
    CityKloster Irsee
    Period7/3/117/8/11

    Fingerprint

    Parameterization
    Diffeomorphisms
    Data-driven
    Control Points
    Template
    Descriptors
    Atlas
    Gradient Descent
    Large Deformation
    Sparsity
    Estimate
    Registration
    Penalty
    Computational Cost
    Finite Set
    Optimization
    Costs

    ASJC Scopus subject areas

    • Computer Science(all)
    • Theoretical Computer Science

    Cite this

    Durrleman, S., Prastawa, M., Gerig, G., & Joshi, S. (2011). Optimal data-driven sparse parameterization of diffeomorphisms for population analysis. In Information Processing in Medical Imaging - 22nd International Conference, IPMI 2011, Proceedings (Vol. 6801 LNCS, pp. 123-134). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 6801 LNCS). https://doi.org/10.1007/978-3-642-22092-0_11

    Optimal data-driven sparse parameterization of diffeomorphisms for population analysis. / Durrleman, Sandy; Prastawa, Marcel; Gerig, Guido; Joshi, Sarang.

    Information Processing in Medical Imaging - 22nd International Conference, IPMI 2011, Proceedings. Vol. 6801 LNCS 2011. p. 123-134 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 6801 LNCS).

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

    Durrleman, S, Prastawa, M, Gerig, G & Joshi, S 2011, Optimal data-driven sparse parameterization of diffeomorphisms for population analysis. in Information Processing in Medical Imaging - 22nd International Conference, IPMI 2011, Proceedings. vol. 6801 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 6801 LNCS, pp. 123-134, 22nd International Conference on Information Processing in Medical Imaging, IPMI 2011, Kloster Irsee, Germany, 7/3/11. https://doi.org/10.1007/978-3-642-22092-0_11
    Durrleman S, Prastawa M, Gerig G, Joshi S. Optimal data-driven sparse parameterization of diffeomorphisms for population analysis. In Information Processing in Medical Imaging - 22nd International Conference, IPMI 2011, Proceedings. Vol. 6801 LNCS. 2011. p. 123-134. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-642-22092-0_11
    Durrleman, Sandy ; Prastawa, Marcel ; Gerig, Guido ; Joshi, Sarang. / Optimal data-driven sparse parameterization of diffeomorphisms for population analysis. Information Processing in Medical Imaging - 22nd International Conference, IPMI 2011, Proceedings. Vol. 6801 LNCS 2011. pp. 123-134 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
    @inproceedings{95adea19e55b49c18c409310747e63cc,
    title = "Optimal data-driven sparse parameterization of diffeomorphisms for population analysis",
    abstract = "In this paper, we propose a novel approach for intensity based atlas construction from a population of anatomical images, that estimates not only a template representative image but also a common optimal parameterization of the anatomical variations evident in the population. First, we introduce a discrete parameterization of large diffeomorphic deformations based on a finite set of control points, so that deformations are characterized by a low dimensional geometric descriptor. Second, we optimally estimate the position of the control points in the template image domain. As a consequence, control points move to where they are needed most to capture the geometric variability evident in the population. Third, the optimal number of control points is estimated by using a log - L1 sparsity penalty. The estimation of the template image, the template-to-subject mappings and their optimal parameterization is done via a single gradient descent optimization, and at the same computational cost as independent template-to-subject registrations. We present results that show that the anatomical variability of the population can be encoded efficiently with these compact and adapted geometric descriptors.",
    author = "Sandy Durrleman and Marcel Prastawa and Guido Gerig and Sarang Joshi",
    year = "2011",
    doi = "10.1007/978-3-642-22092-0_11",
    language = "English (US)",
    isbn = "9783642220913",
    volume = "6801 LNCS",
    series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
    pages = "123--134",
    booktitle = "Information Processing in Medical Imaging - 22nd International Conference, IPMI 2011, Proceedings",

    }

    TY - GEN

    T1 - Optimal data-driven sparse parameterization of diffeomorphisms for population analysis

    AU - Durrleman, Sandy

    AU - Prastawa, Marcel

    AU - Gerig, Guido

    AU - Joshi, Sarang

    PY - 2011

    Y1 - 2011

    N2 - In this paper, we propose a novel approach for intensity based atlas construction from a population of anatomical images, that estimates not only a template representative image but also a common optimal parameterization of the anatomical variations evident in the population. First, we introduce a discrete parameterization of large diffeomorphic deformations based on a finite set of control points, so that deformations are characterized by a low dimensional geometric descriptor. Second, we optimally estimate the position of the control points in the template image domain. As a consequence, control points move to where they are needed most to capture the geometric variability evident in the population. Third, the optimal number of control points is estimated by using a log - L1 sparsity penalty. The estimation of the template image, the template-to-subject mappings and their optimal parameterization is done via a single gradient descent optimization, and at the same computational cost as independent template-to-subject registrations. We present results that show that the anatomical variability of the population can be encoded efficiently with these compact and adapted geometric descriptors.

    AB - In this paper, we propose a novel approach for intensity based atlas construction from a population of anatomical images, that estimates not only a template representative image but also a common optimal parameterization of the anatomical variations evident in the population. First, we introduce a discrete parameterization of large diffeomorphic deformations based on a finite set of control points, so that deformations are characterized by a low dimensional geometric descriptor. Second, we optimally estimate the position of the control points in the template image domain. As a consequence, control points move to where they are needed most to capture the geometric variability evident in the population. Third, the optimal number of control points is estimated by using a log - L1 sparsity penalty. The estimation of the template image, the template-to-subject mappings and their optimal parameterization is done via a single gradient descent optimization, and at the same computational cost as independent template-to-subject registrations. We present results that show that the anatomical variability of the population can be encoded efficiently with these compact and adapted geometric descriptors.

    UR - http://www.scopus.com/inward/record.url?scp=79959626160&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=79959626160&partnerID=8YFLogxK

    U2 - 10.1007/978-3-642-22092-0_11

    DO - 10.1007/978-3-642-22092-0_11

    M3 - Conference contribution

    C2 - 21761651

    AN - SCOPUS:79959626160

    SN - 9783642220913

    VL - 6801 LNCS

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

    SP - 123

    EP - 134

    BT - Information Processing in Medical Imaging - 22nd International Conference, IPMI 2011, Proceedings

    ER -