Blog
About

21
views
0
recommends
+1 Recommend
0
shares
  • Review: found
Is Open Access

Review of 'Intersubject information mapping: revealing canonical representations of complex natural stimuli'

Bookmark
5
Very interesting proposal for data analysis of fMRI acquired using naturalistic stimuli
Average rating:
    Rated 4.5 of 5.
Level of importance:
    Rated 5 of 5.
Level of validity:
    Rated 4 of 5.
Level of completeness:
    Rated 4 of 5.
Level of comprehensibility:
    Rated 5 of 5.
Competing interests:
None

Reviewed article

  • Record: found
  • Abstract: found
  • Article: found
Is Open Access

Intersubject information mapping: revealing canonical representations of complex natural stimuli

 Nikolaus Kriegeskorte (corresponding) (2015)
Abstract Real-world time-continuous stimuli such as video promise greater naturalism for studies of brain function. However, modeling the stimulus variation is challenging and introduces a bias in favor of particular descriptive dimensions. Alternatively, we can look for brain regions whose signal is correlated between subjects, essentially using one subject to model another. Intersubject correlation mapping (ICM) allows us to find brain regions driven in a canonical manner across subjects by a complex natural stimulus. However, it requires a direct voxel-to-voxel match between the spatiotemporal activity patterns and is thus only sensitive to common activations sufficiently extended to match up in Talairach space (or in an alternative, e.g. cortical-surface-based, common brain space). Here we introduce the more general approach of intersubject information mapping (IIM). For each brain region, IIM determines how much information is shared between the subjects' local spatiotemporal activity patterns. We estimate the intersubject mutual information using canonical correlation analysis applied to voxels within a spherical searchlight centered on each voxel in turn. The intersubject information estimate is invariant to linear transforms including spatial rearrangement of the voxels within the searchlight. This invariance to local encoding will be crucial in exploring fine-grained brain representations, which cannot be matched up in a common space and, more fundamentally, might be unique to each individual – like fingerprints. IIM yields a continuous brain map, which reflects intersubject information in fine-grained patterns. Performed on data from functional magnetic resonance imaging (fMRI) of subjects viewing the same television show, IIM and ICM both highlighted sensory representations, including primary visual and auditory cortices. However, IIM revealed additional regions in higher association cortices, namely temporal pole and orbitofrontal cortex. These regions appear to encode the same information across subjects in their fine-grained patterns, although their spatial-average activation was not significantly correlated between subjects.
    Bookmark

    Review information

    10.14293/S2199-1006.1.SOR-SOCSCI.APDIXF.v1.RXSXYL

    This work has been published open access under Creative Commons Attribution License CC BY 4.0, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Conditions, terms of use and publishing policy can be found at www.scienceopen.com.

    ScienceOpen disciplines:
    Keywords:

    Review text

    As the paper states in its introduction, one important way towards imaging studies of the behavior of the brain under natural conditions is the use of time continuous natural stimuli such as videos. However, this poses a data-analysis challenge since fully modelling such multidimensional stimuli is impossible. Inter-subject correlation (ISC, or intersubject correlation mapping (ICM)) is an already established technique of analysis of naturalistic stimulus data in fMRI (Hasson 2004). In this technique, subject’s bold responses are used to model other subjects’ bold responses. This ISC/ICM technique despite of its apparent simplicity has been shown to be surprisingly sensitive of detecting activations also with standard block design stimuli albeit the technique does not utilize any model of the stimulus (Pajula 2012). However, as the paper under review states, ISC mapping requires spatial matching of the subjects in Talairach space and therefore differences in individual functional anatomy (and the limited registration accuracy) limit its sensitivity particularly when studying higher order brain functions. This sensitivity problem is serious, I argue that it is more serious than in the traditional GLM based fMRI analysis, and often addressed by (excessive) smoothing. The paper under review proposes a much more sophisticated approach to solve the problem. Namely, it proposes to compute the first canonical correlations within search-light regions around each voxel. To me, this appears as very interesting idea and first results using this approach look promising. I only have few remarks about the paper, answering to which could further improve this very interesting work.




    • I somehow managed to miss the number of subjects participating in the experiment.

    • The paper does not make it clear if inter-subject information is measured between two subjects instead of n > 2 subjects . If it is measured between only two subjects, it would be interesting to obtain some insights how to do the analysis if there is a larger set of subjects. There are at least a few ways to combine pairwise inter subject correlation values (Hanson 2009, Lerner 2011, Kauppi 2014) to form a group statistic between n subjects. It would be interesting to obtain insights do these methods generalize to IIM.

    • Although the connection of canonical correlation to mutual information is interesting, in addition it would be illuminating to mention that the first canonical correlation is equivalent to maximum correlation coefficient between linear combinations of time courses (each normalized to have unit variances) within the two search-lights (Johnson, Wichern 1992).

    • I missed a figure about ICM results corresponding to IIM presented in figure 5.



    References:



    Hanson SJ, Gagliardi AD, Hanson C (2009) Solving the brain synchrony eigenvalue problem: Conservation of temporal dynamics (fmri) over subjects doing the same task. Journal of computational neuroscience 27: 103–114.



    Hasson U, Nir Y, Levy I, Fuhrmann G, Malach R (2004) Intersubject synchronization of cortical activity during natural vision. Science 303: 1634–1640



    Johnson, R. A., & Wichern, D. W. (1992). Applied multivariate statistical analysis. Englewood Cliffs, NJ: Prentice hall.



    Kauppi JP, Pajula J, Tohka J. A versatile software package for inter-subject correlation based analyses of fMRI. Front Neuroinform. 2014 Jan 31;8:2.



    Lerner, Y., Honey, C. J., Silbert, L. J., and Hasson, U. (2011). Topographic mapping of a hierarchy of temporal receptive windows using a narrated story. J. Neurosci. 31, 2906–2915.



    Pajula, J., Kauppi, J.-P., and Tohka, J. (2012). Inter-subject correlation in fMRI: Method validation against stimulus-model based analysis. PLoS ONE 7:e41196.



    Comments

    Comment on this review