Get access

Large sample group independent component analysis of functional magnetic resonance imaging using anatomical atlas-based reduction and bootstrapped clustering

Authors

  • Ariana Anderson,

    Corresponding author
    1. Department of Psychiatry and Biobehavioral Sciences, University of California at Los Angeles, Los Angeles, CA
    • Department of Psychiatry and Biobehavioral Sciences, University of California at Los Angeles, Los Angeles, CA
    Search for more papers by this author
  • Jennifer Bramen,

    1. Department of Psychiatry and Biobehavioral Sciences, University of California at Los Angeles, Los Angeles, CA
    Search for more papers by this author
  • Pamela K. Douglas,

    1. Department of Psychiatry and Biobehavioral Sciences, University of California at Los Angeles, Los Angeles, CA
    Search for more papers by this author
  • Agatha Lenartowicz,

    1. Department of Psychiatry and Biobehavioral Sciences, University of California at Los Angeles, Los Angeles, CA
    Search for more papers by this author
  • Andrew Cho,

    1. Department of Psychiatry and Biobehavioral Sciences, University of California at Los Angeles, Los Angeles, CA
    Search for more papers by this author
  • Chris Culbertson,

    1. Department of Psychiatry and Biobehavioral Sciences, University of California at Los Angeles, Los Angeles, CA
    Search for more papers by this author
  • Arthur L. Brody,

    1. Department of Psychiatry and Biobehavioral Sciences, University of California at Los Angeles, Los Angeles, CA
    2. Department of Psychiatry, Greater Los Angeles VA Healthcare System, Los Angeles, CA
    3. Department of Radiology, Greater Los Angeles VA Healthcare System, Los Angeles, CA
    Search for more papers by this author
  • Alan L. Yuille,

    1. Department of Statistics, University of California at Los Angeles, Los Angeles, CA
    2. Department of Brain and Cognitive Engineering, Korea University, Anam-dong, Seongbuk-gu, Seoul 136-713, Korea
    Search for more papers by this author
  • Mark S. Cohen

    1. Department of Psychiatry and Biobehavioral Sciences, University of California at Los Angeles, Los Angeles, CA
    2. Department of Biomedical Engineering, Neurology, Psychology, Biomedical Physics and Radiology, University of California at Los Angeles, Los Angeles, CA
    Search for more papers by this author

  • This work is supported by the National Institute on Drug Abuse under DA026109 to M.S.C and by WCU (World Class University) program through the National Research Foundation of Korea funded by the Ministry of Education, Science and Technology (R31-2008-000-10008-0) to AY. The grant funding that supported the data collection was: the National Institute on Drug Abuse (A.L.B. [R01 DA20872]), a Veterans Affairs Type I Merit Review Award (A.L.B.), and an endowment from the Richard Metzner Chair in Clinical Neuropharmacology (A.L.B.). We thank Michael Durnhofer for maintaining the systems necessary for these data analyses.

Abstract

Independent component analysis (ICA) is a popular method for the analysis of functional magnetic resonance imaging (fMRI) signals that is capable of revealing connected brain systems of functional significance. To be computationally tractable, estimating the independent components (ICs) inevitably requires one or more dimension reduction steps. Whereas most algorithms perform such reductions in the time domain, the input data are much more extensive in the spatial domain, and there is broad consensus that the brain obeys rules of localization of function into regions that are smaller in number than the number of voxels in a brain image. These functional units apparently reorganize dynamically into networks under different task conditions. Here we develop a new approach to ICA, producing group results by bagging and clustering over hundreds of pooled single-subject ICA results that have been projected to a lower-dimensional subspace. Averages of anatomically based regions are used to compress the single subject-ICA results prior to clustering and resampling via bagging. The computational advantages of this approach make it possible to perform group-level analyses on datasets consisting of hundreds of scan sessions by combining the results of within-subject analysis, while retaining the theoretical advantage of mimicking what is known of the functional organization of the brain. The result is a compact set of spatial activity patterns that are common and stable across scan sessions and across individuals. Such representations may be used in the context of statistical pattern recognition supporting real-time state classification. © 2011 Wiley Periodicals, Inc. Int J Imaging Syst Technol, 21, 223–231, 2011

Ancillary