Bayesian covariate selection in mixed-effects models for longitudinal shape analysis

Prasanna Muralidharan, James Fishbaugh, Eun Young Kim, Hans J. Johnson, Jane S. Paulsen, Guido Gerig, P. Thomas Fletcher

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

    Abstract

    The goal of longitudinal shape analysis is to understand how anatomical shape changes over time, in response to biological processes, including growth, aging, or disease. In many imaging studies, it is also critical to understand how these shape changes are affected by other factors, such as sex, disease diagnosis, IQ, etc. Current approaches to longitudinal shape analysis have focused on modeling age-related shape changes, but have not included the ability to handle covariates. In this paper, we present a novel Bayesian mixed-effects shape model that incorporates simultaneous relationships between longitudinal shape data and multiple predictors or covariates to the model. Moreover, we place an Automatic Relevance Determination (ARD) prior on the parameters, that lets us automatically select which covariates are most relevant to the model based on observed data. We evaluate our proposed model and inference procedure on a longitudinal study of Huntington's disease from PREDICT-HD. We first show the utility of the ARD prior for model selection in a univariate modeling of striatal volume, and next we apply the full high-dimensional longitudinal shape model to putamen shapes.

    Original languageEnglish (US)
    Title of host publication2016 IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2016 - Proceedings
    PublisherIEEE Computer Society
    Pages656-659
    Number of pages4
    Volume2016-June
    ISBN (Electronic)9781479923502
    DOIs
    StatePublished - Jun 15 2016
    Event2016 IEEE 13th International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2016 - Prague, Czech Republic
    Duration: Apr 13 2016Apr 16 2016

    Other

    Other2016 IEEE 13th International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2016
    CountryCzech Republic
    CityPrague
    Period4/13/164/16/16

    Fingerprint

    Biological Phenomena
    Corpus Striatum
    Aptitude
    Putamen
    Huntington Disease
    Longitudinal Studies
    Growth
    Aging of materials
    Imaging techniques

    Keywords

    • Bayesian analysis
    • Huntington's disease
    • Longitudinal shape analysis
    • model selection

    ASJC Scopus subject areas

    • Biomedical Engineering
    • Radiology Nuclear Medicine and imaging

    Cite this

    Muralidharan, P., Fishbaugh, J., Kim, E. Y., Johnson, H. J., Paulsen, J. S., Gerig, G., & Fletcher, P. T. (2016). Bayesian covariate selection in mixed-effects models for longitudinal shape analysis. In 2016 IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2016 - Proceedings (Vol. 2016-June, pp. 656-659). [7493352] IEEE Computer Society. https://doi.org/10.1109/ISBI.2016.7493352

    Bayesian covariate selection in mixed-effects models for longitudinal shape analysis. / Muralidharan, Prasanna; Fishbaugh, James; Kim, Eun Young; Johnson, Hans J.; Paulsen, Jane S.; Gerig, Guido; Fletcher, P. Thomas.

    2016 IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2016 - Proceedings. Vol. 2016-June IEEE Computer Society, 2016. p. 656-659 7493352.

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

    Muralidharan, P, Fishbaugh, J, Kim, EY, Johnson, HJ, Paulsen, JS, Gerig, G & Fletcher, PT 2016, Bayesian covariate selection in mixed-effects models for longitudinal shape analysis. in 2016 IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2016 - Proceedings. vol. 2016-June, 7493352, IEEE Computer Society, pp. 656-659, 2016 IEEE 13th International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2016, Prague, Czech Republic, 4/13/16. https://doi.org/10.1109/ISBI.2016.7493352
    Muralidharan P, Fishbaugh J, Kim EY, Johnson HJ, Paulsen JS, Gerig G et al. Bayesian covariate selection in mixed-effects models for longitudinal shape analysis. In 2016 IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2016 - Proceedings. Vol. 2016-June. IEEE Computer Society. 2016. p. 656-659. 7493352 https://doi.org/10.1109/ISBI.2016.7493352
    Muralidharan, Prasanna ; Fishbaugh, James ; Kim, Eun Young ; Johnson, Hans J. ; Paulsen, Jane S. ; Gerig, Guido ; Fletcher, P. Thomas. / Bayesian covariate selection in mixed-effects models for longitudinal shape analysis. 2016 IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2016 - Proceedings. Vol. 2016-June IEEE Computer Society, 2016. pp. 656-659
    @inproceedings{fb33f082082f4b43ae9d21866381e2ad,
    title = "Bayesian covariate selection in mixed-effects models for longitudinal shape analysis",
    abstract = "The goal of longitudinal shape analysis is to understand how anatomical shape changes over time, in response to biological processes, including growth, aging, or disease. In many imaging studies, it is also critical to understand how these shape changes are affected by other factors, such as sex, disease diagnosis, IQ, etc. Current approaches to longitudinal shape analysis have focused on modeling age-related shape changes, but have not included the ability to handle covariates. In this paper, we present a novel Bayesian mixed-effects shape model that incorporates simultaneous relationships between longitudinal shape data and multiple predictors or covariates to the model. Moreover, we place an Automatic Relevance Determination (ARD) prior on the parameters, that lets us automatically select which covariates are most relevant to the model based on observed data. We evaluate our proposed model and inference procedure on a longitudinal study of Huntington's disease from PREDICT-HD. We first show the utility of the ARD prior for model selection in a univariate modeling of striatal volume, and next we apply the full high-dimensional longitudinal shape model to putamen shapes.",
    keywords = "Bayesian analysis, Huntington's disease, Longitudinal shape analysis, model selection",
    author = "Prasanna Muralidharan and James Fishbaugh and Kim, {Eun Young} and Johnson, {Hans J.} and Paulsen, {Jane S.} and Guido Gerig and Fletcher, {P. Thomas}",
    year = "2016",
    month = "6",
    day = "15",
    doi = "10.1109/ISBI.2016.7493352",
    language = "English (US)",
    volume = "2016-June",
    pages = "656--659",
    booktitle = "2016 IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2016 - Proceedings",
    publisher = "IEEE Computer Society",
    address = "United States",

    }

    TY - GEN

    T1 - Bayesian covariate selection in mixed-effects models for longitudinal shape analysis

    AU - Muralidharan, Prasanna

    AU - Fishbaugh, James

    AU - Kim, Eun Young

    AU - Johnson, Hans J.

    AU - Paulsen, Jane S.

    AU - Gerig, Guido

    AU - Fletcher, P. Thomas

    PY - 2016/6/15

    Y1 - 2016/6/15

    N2 - The goal of longitudinal shape analysis is to understand how anatomical shape changes over time, in response to biological processes, including growth, aging, or disease. In many imaging studies, it is also critical to understand how these shape changes are affected by other factors, such as sex, disease diagnosis, IQ, etc. Current approaches to longitudinal shape analysis have focused on modeling age-related shape changes, but have not included the ability to handle covariates. In this paper, we present a novel Bayesian mixed-effects shape model that incorporates simultaneous relationships between longitudinal shape data and multiple predictors or covariates to the model. Moreover, we place an Automatic Relevance Determination (ARD) prior on the parameters, that lets us automatically select which covariates are most relevant to the model based on observed data. We evaluate our proposed model and inference procedure on a longitudinal study of Huntington's disease from PREDICT-HD. We first show the utility of the ARD prior for model selection in a univariate modeling of striatal volume, and next we apply the full high-dimensional longitudinal shape model to putamen shapes.

    AB - The goal of longitudinal shape analysis is to understand how anatomical shape changes over time, in response to biological processes, including growth, aging, or disease. In many imaging studies, it is also critical to understand how these shape changes are affected by other factors, such as sex, disease diagnosis, IQ, etc. Current approaches to longitudinal shape analysis have focused on modeling age-related shape changes, but have not included the ability to handle covariates. In this paper, we present a novel Bayesian mixed-effects shape model that incorporates simultaneous relationships between longitudinal shape data and multiple predictors or covariates to the model. Moreover, we place an Automatic Relevance Determination (ARD) prior on the parameters, that lets us automatically select which covariates are most relevant to the model based on observed data. We evaluate our proposed model and inference procedure on a longitudinal study of Huntington's disease from PREDICT-HD. We first show the utility of the ARD prior for model selection in a univariate modeling of striatal volume, and next we apply the full high-dimensional longitudinal shape model to putamen shapes.

    KW - Bayesian analysis

    KW - Huntington's disease

    KW - Longitudinal shape analysis

    KW - model selection

    UR - http://www.scopus.com/inward/record.url?scp=84978388458&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=84978388458&partnerID=8YFLogxK

    U2 - 10.1109/ISBI.2016.7493352

    DO - 10.1109/ISBI.2016.7493352

    M3 - Conference contribution

    VL - 2016-June

    SP - 656

    EP - 659

    BT - 2016 IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2016 - Proceedings

    PB - IEEE Computer Society

    ER -