CIFTI file is a file format that stores data from surfaces (vertices)
and volumes (voxels) concurrently in a single file comprising a listed
set of brainordinates (brain coordinates). The volume component of a
CIFTI file can be any selected list of voxels (e.g., only subcortical
gray-matter voxels) that need not conform to cuboidal or ‘N x M x O’
The Connectome Workbench currently supports CIFTI files for dense time
series (.dtseries.nii), dense connectomes (.dconn.nii), dense labels (*.dlabel.nii)
- dense conectomes CIFTI (connectivity matrix) refers to connectivity
matrix CIFTI files that contain correlations between every surface or
volume brainordinate and every other surface or volume ordinate in our
model of the brain. For example, a grayordinates × grayordinates dense
connectome file contains a connectivity value between every
grayordinate and every other grayordinate. Dense connectomes can also
be asymmetric, such as a structural connectivity matrix between
grayordinates (all gray matter structures) and whiteordinates (all
white matter structures).
brainordinates x brainordinates
- dense connectome labels CIFTI files (grayordinates × label maps)
contain a name and/or color label for every brainordinate. These files
will be useful for mapping combined cortical and subcortical
parcellations. dense scalar CIFTI files contain real-valued data for
every brainordinate mapped to a color palette. A dense scalar file can
contain one map or frame (such as a single cortical myelin map) or a
series of spatial maps (e.g. tfMRI contrast maps or ICA component
spatial maps) that can have distinct labels for each frame.
- dense time-series is a type of dense scalar CIFTI file
(grayordinates × time) that can be viewed as a series of maps (or
animated movie) displayed on cortical surfaces and subcortical
volume slices or as a timecourse graph for a selected grayordinate,
using the timecourse graph window (and as an average timecourse for
all grayordinates within a selected ROI). This type of file is
particularly useful for viewing resting state fMRI time series.
(brain oridnates x time points)
What we ultimately want is a dense connectome matrix based
on/resulting from a tractography method from diffusion data. What is
the relationship between the coordinates of diffusion voxels and the
node indices of the surfaces of structural images?
The CIFTI standard space is in MNI space (subcortical voxels). The
correspondgin surface is also in MNI space. From some discussion I
read that the cifti files do not contain surface coordinates.