|
|
Comparing functional and cytoarchitectonic data of human V5/MT+
|
Marcus Wilms1, Alexander Malikovic1, Simon Eickhoff1, Karsten Specht1,3, Hartmut Mohlberg1, Katrin Amunts1, Gereon Fink1,2,3
1Institute of Medicine, Research Centre Jülich, Germany, 2Brain Imaging Centre West, Research Centre Jülich, Germany, 3Department of Neurology – Cognitive Neurology, University Hospital, Aachen, Germany
|
Objective:
Thus far the delineation of the human visual “motion area” relied
mainly on functional data using paradigms inspired by monkey studies of
area MT. We here compare functional magnetic resonance imaging
(fMRI) data from normal volunteers of a functionally identified motion
sensitive visual area with its cytoarchitectonic correlate derived from
microscopic studies of postmortem brains.
Methods: Using fMRI (Siemens SONATA, 1.5T), putative human area V5/MT+
was functionally identified (n=14, 5M+9F, 23±5 yr) by modelling the
BOLD responses to alternating radially moving and stationary dot
patterns similar to those used in other studies [1,2]. Single subject t-statistic contrasts for dynamic vs. stationary
stimulation were entered into a random-effects analysis comparing the
mean activation to the variability in activations from subject to
subject (small volume correction for the a priori predicted V5/MT+
region based on literature data). In addition, we calculated functional
probability maps (fPM) by thresholding the individual t-maps at pcorr<0.01 and setting all significant voxels to unity, non-significant voxels to zero. These binarized t-maps
were then added over all subjects on a voxel-by-voxel basis.
Observer-independent cytoarchitectonic mapping data was obtained from
10 human postmortem brains [3]. The superimposition of individual
cytoarchitectonic maps yielded anatomical probability maps (aPM) much
alike the fPMs [4]. Functionally and cytoarchitectonically defined
putative human equivalents of area V5/MT+ were then compared after spatial normalization to the single-subject MNI reference brain [5].
Results & Discussion:
Bilateral visual cortex activations were seen in the single subject dynamic vs. stationary functional contrasts. Significant bilateral activations were also seen in the predicted region of V5/MT+
in the group level random-effects analysis. Comparison of this group
data with the aPMs revealed that 14.7 % (15.9 %) of the right (left)
functional activation were assigned to the right (left) aPM. Reversely,
97.0 % (55.1 %) of the right (left) functional activation were covered
by the respective aPM. Comparison of functional with anatomical PMs
showed that 34.7 % (24.2 %) of the right (left) fPM were assigned to
the right (left) aPM. In turn, 87.1 % (62.3 %) of the right (left) aPM
were covered by the respective fPM.
Conclusions:
Random-effects data as well as fPMs yield similar results. Due to the
method of their generation, fPMs more closely resemble the aPMs. The
convergence of both data modalities shows that the functionally defined
human V5/MT+ correlates with the cytoarchitectonic V5/MT+ map obtained from postmortem brains.
References & Acknowledgements:
[1] Huk et al. 2002. J Neuroscience, 22:7195-7205
[2] Tootell et al. 1995. J Neuroscience, 15,4:3215-3230
[3] Malikovic et al. 2001. HBM Conference 2001
[4] Eickhoff et al. 2005. HBM Conference 2005
[5] Eickhoff et al. 2005. NeuroImage, (in press)
|
|
|