3D Probabilistic map of cytoarchitectonically defined human thalamic nuclei in standard MRI

Abstract

Event Abstract Back to Event 3D Probabilistic map of cytoarchitectonically defined human thalamic nuclei in standard MRI Harry B. Uylings1*, Katrin Amunts2, H. Mohlberg3, C. W. Pool1, K. Zilles3 and W. J. Smeets4 1 VU Univ. Med. Ctr./ Univ.Maastricht, Netherlands 2 Research Ctr Juelich, RWTH Aachen Univ, Netherlands 3 Research Ctr Juelich, GERMANY, Germany 4 Netherlands Institute for Brain Research, Netherlands Introduction: The parcellation of the human thalamus has been the subject of several studies (Morel et al., 1997; Jones, 2006). Apart from geometric variability of macroscopic anatomical features, there is a very large variability in microscopically defined brain areas even after the application of the full non-linear deformation method (Amunts et al., 1999; Uylings et al, 2005). For the interpretation of functional imaging data it is essential to know the extent of interindividual variability of the different thalamic nuclei. Some indications for variability in thalamic nuclei are reported by Niemann et al., 2000, and Johansen-Berg et al, 2005. Data on the variability in cytoarchitectonically defined thalamic nuclei displayed in the stereotactic format of the MNI brain are here provided for the first time. Methods: The thalamus of ten control human brains (5 males) has been studied. Postmortem MRIs , scanned in a 1.5 T Siemens Magnetom, were taken before the brains were embedded, cut in the frontal plane and stained for cell bodies. Cytoarchitectonic criteria as cell density, cell size and shape, and arrangement of cells were used to delineate about 30 thalamic nuclei in these sections taken at an interval of about 1 mm. The cytoarchitectonic delineations were 3D-reconstructed and then transformed to the reference MNI brain by elastic deformation. 3D probabilistic maps were calculated which describe, for each voxel of the reference space, the relative frequency with which a nucleus was present at this particular position. Results and Conclusion: The thalamic nuclei show a smaller inter-individual variability after transformation into MNI space than the cortical areas studied. Still the inter-individual variability in cytoarchitectonically defined thalamic nuclei is considerable. Intersubject variability differs between nuclei within the thalamus. The mediodorsal nucleus (MD), e.g., Fig.1, has a low variability, whereas the variability of ventral posterior lateral nucleus (VPL), Fig2, and medial geniculate nucleus (MGN), Fig3, is higher. These maps will open new perspective for the interpretation of functional imaging data for thalamic structural-functional analyses in the human brain, where many techniques are not available.