The International Neuroimaging Consortium (INC) is an international multi-center consortium with collaborators in the United States, Denmark and Japan supported by the Human Brain Project. One of the goals of this consortium is to create modeling, analysis and visualization tools which assist in the determination and quantification of spatial and temporal patterns in functional neuroimaging datasets.
CirclePack
CirclePack is a software package which can be used to
create flat maps of the brain. We use a novel circle packing algorithm
which produces maps that exhibit conformal behavior and are mathematically
unique. These maps are quasi-conformal, as angular proportion between the
original surface and the flattened surface is controlled, and there is
a bounded amount of angular distortion.
This approach offers several advantages over existing approaches. We can produces maps in the conventional Euclidean plane and on a sphere. We can also produce maps in the hyperbolic plane, which offers the advantage of being able to change the map focus interactively and alter the locations of map distortion. In addition, these maps are unique and a coordinate system can easily be imposed on the maps by specifying two anatomical landmarks; color can be used to identify anatomical and functional regions of interest. This software is demonstrated, and map results which use functional and anatomical data from human cortical surfaces are presented.
Corner Cube
Corner Cube is a software package that can be used to
extract activated foci from neuroimaging datasets and visualize them as
symbolic representations in a three-dimensional anatomical context.
Foci are extracted by thresholding a data volume and identifying the distinct
clusters that remain; statistical metrics are calculated and Talairach
locations and anatomical labelling are reported for each focus. Alternatively,
foci that have been reported in the literature using Talairach millimetric
coordinates can be imported to Corner Cube via a few lines of formatted
text. The foci are represented by symbols that preserve the centroid
and volume of the clustered voxels; symbols can vary in complexity from
spheres to surface renderings of the clusters. Anatomical context
is provided by presenting MRI images on three walls (a corner) of the cube
bounding the dataset and projecting the outlines of the foci onto the walls.
The advantages of the approach are: an entire dataset can be summarized in a single picture for a quick overview of the relative size and position of activated foci, two datasets can be represented simultaneously, and datasets described in the literature can be easily imported for comparison to each other.
This work is supported in part by NIH grants MH57180 and NS33718.