An attempt for coloring multichannel MR imaging data

This is an elementary research into assigning color values to voxels of multi-channel magnetic resonance imaging (MRI) volume data. The MRI volume data sets obtained under different scanning conditions are transformed into components by independent component analysis (ICA), which enhances the physical characteristics of the tissue. The transfer functions for generating color values from the independent components are obtained by using a radial basis function network, a kind of neural net, by training the network with sample data chosen from the Visible Human female data set (VHF). The resultant color volume data sets correspond well with the full-color cross-sections of the Visible Human data sets