Elsevier

NeuroImage

Volume 47, Supplement 1, July 2009, Page S162
NeuroImage

Informatics and imaging pipeline for development of a new multi-modal atlas of the human brain: imaging, anatomy, and gene expression

https://doi.org/10.1016/S1053-8119(09)71707-0Get rights and content

Introduction

Classically, cytoarchitectonic maps derived from the investigations of Brodmann (1909) and von Economo (1925) have been used to define regions of functional specialization (Eickhoff et al., 2007). This analysis is typically done at the exclusion of the ensemble of cyto-, myelo-, chemo- and gene expression architectures defining the human brain (Toga et al., 2006; Nowinski 2007). Inspired by our previous work on the mouse brain (Lein et al., 2007) and human cortex (www.humancortex.alleninstitute.org), the Allen Institute for Brain Science has begun a four-year project to develop a multi-modal and multi-resolution human brain atlas comprising 3D magnetic resonance imaging (MRI), detailed histology, and spatially mapped gene expression data for several human brains. The resulting atlas will be a free publicly available resource and will allow users to assess gene expression in its anatomic and cytoarchitectonic context, thus allowing researchers to connect anatomic and functional information with underlying genetic information. At the core of this atlas is the integration of the varied data modalities within a unified spatial framework based on the volumetric MRI data.

Section snippets

Methods

Volumetric T1- and T2-weighted MRIs were along with diffusion tensor imaging hyper-volumes (32 different directions for ex cranio data and 64 for in cranio) were acquired at 3 Tesla. Each brain was then separated into serial 5 mm coronal slabs and blockfaces were photographically imaged. Full coronal sections (50 μm) were histochemically stained with Hemotoxylin and Eosin, thionin (Nissl stain), Bielschowski silver stain for axons, and immunohistochemically with SMI-32 to visualize pyramidal

Results

The results demonstrate how the volumetric images can be used as a 3D scaffold for the multi-modal data generated (including microarray-based gene expression). We further demonstrate how experimental imaging data may be customized to our atlas to evaluate relevant gene expression patterns.

Conclusions

Our work demonstrates the next phase of exploration in human brain atlasing efforts and the integration of classical histology, immunohistochemistry, functional imaging, and gene expression analysis. Though only correspondence with MRI-based data is demonstrated here, alignment with other functional modalities such as positron emission tomography could be implemented. The atlas that we are generating will, for the first time, allow different groups to simultaneously examine the gene expression,

References (0)

Cited by (0)

View full text