Diffusion Tensor Imaging Analysis

The Diffusion Analysis dialog contains the following tabs

• Diffusion Weighted Imaging
• Diffusion Tensor Imaging

Note: The Diffusion Tensor Imaging tab can be displayed by selecting Diffusion Analysis (DTI) on the Modules menu.

Acquisition parameters:

Number of gradient directions: Here the number of gradient directions in the input images is displayed. If this information is not yet defined, the value will display here. In this case manual definition is required by clicking the Edit configuration button.

Number of b0 images per slice: Here the number of b=0 (unweighted) images per slice in the input images is displayed. If this information is not yet defined, the value will display Not specified here. In this case manual definition is required by clicking the Edit configuration button.

b-Value: The b-value of the diffusion weighted images.

Edit configuration: Opens the manual gradient configuration dialog box.

Current configuration: Indicates which configuration (diffusion directions and b-value) are currently active. When this field reads "Extracted from header" it means that the above settings are read from the DICOM header of the input images. Alternatively, this may specify a previously user-defined configuration set in the Edit configuration dialog box.

Preprocessing:

Auto detect noise threshold: Toggle on/off auto-detection of noise threshold.

Show noise level cutoff: Show the pixels in the input images that are below current Noise level threshold in red.

Noise level: Disregard all pixels below this threshold from the analysis.

Average images:  This option will be available if either multiple b=0 images per slice are present, and/or multiple repeated gradient configurations was detected in the input images. If enabled, the repeated images will be averaged prior to the analysis.

Smooth images: Apply nearest neighbour smoothing to the input data before analysis. 

Motion correction: When selected, motion correction will be performed prior to diffusion map generation. See Motion and eddy current correction for details on the algorithms used.

Eddy current correction: When selected, eddy current compensation will be performed prior to diffusion map generation. See Motion and eddy current correction for details on the algorithms used.

Show graphs: Shows the result from the Motion and/or Eddy current correction (if/when performed).

Output

Eigenvector color map (cDTI) : Generates diffusion tensor colormaps (color DTI) where the pixel color reflects the direction of the diffusion tensor in that voxel. The color coding is implemented by scaling the directions of the eigenvector with the largest eigenvalue according to the scheme (AXIAL):

• Red: Anterior-posterior
• Green: Left-right
• Blue: Inferior-superior.

The intensity of the colors are scaled with the Fractional Anisotropy (FA) index. The maximum scaling FA-value to be applied in this scaling can be modified by the Max FA-scale box. Default, the intensity in these maps will be scaled with the FA-value from 0 to 1, but the maximum value can be set e.g. to enhance intensity within regions/images having low FA-values. Note that this factor does not influence the FA-values themselves (FA-maps), but only the visual representation of the cDTI maps on screen. 

Fractional Anisotropy Index (FA) : Generates maps reflecting the fractional anisotropy of each voxel. The index range is 0-1.  The values are calculated according the following formula:

Trace Weighted (TraceW) : Generates maps reflecting the geometric average of the individual diffusion weighted images. These maps are comparable to orthogonal DWI, and has a contrast reversal as compared to the ADC image. 

Mean diffusivity (ADC) : Generates maps reflecting the trace of the diffusion tensor (Tr(D)/3). The mean diffusivity map is scaled so that the pixels are in units of 10^-5 mm² /s.  The values are calculated according to the following formula:

Tensor eigenvalues: Generates (3) maps reflecting the individual eigenvalues of the diffusion tensor. The eigenvalues are sorted in descending order, i.e. lambda_1 > lambda_2 > lambda_3.

Fiber Tracking:

Disable: Check this checkbox if you do not want to perform any fiber tracking (only calculate maps). Disabling the fiber tracking will save resources (memory).

Tracking method: Choose the tracking method approach:

• Search from seed: Tracking is initiated from defined seed-voxels. In this case, the seed-region has to be defined within the image volume using the ROI/VOI feature in the DTI 3D viewer window (see DTI interaction panel in the MPR viewer ). In order to use this method you need to define at least one ROI/VOI that will be used as a seed region. Optionally one or more additional  ROI/VOIs can be defined requiring or excluding fiber passage in certain regions of the brain.

• Exhaustive search: Tracking is initiated from all voxels within the volume (every voxel is defined as seed), and tracts are followed until one of the termination criteria are met. Optionally, up to three different ROIs can be drawn within the image volume in order to select only fibers that pass through the given regions of interest.

Termination criteria: Set the tracking termination criteria:

• Fractional Anisotropy: If the Fractional Anisotropy index of the voxel being propagated to is less than this threshold tracking of the current tract is terminated.

• Tract turning angle: If the angle (in degrees) between two successive principal eigenvectors is larger than this threshold tracking of the current tract is terminated.

• Minimum fiber length: Threshold on the minimum length (in mm) of a reconstructed fiber tract to be kept. This option can be used to suppress very small spurious tracts.

Seeds per voxel: Number of randomly distributed seed-points to be used for every seed-voxel

Perform Fiber Tracking: Select this option to perform Fiber Tracking immediately after DTI maps has been generated. This will cause Fiber Tracking to be performed using an exhaustive search algorithm trying to reconstruct all white matter fiber tracts in the brain.

Process:

Start processing the data

Launch 3D viewer:

Launch the MPR window with the DTI 3D window and DTI Interaction Panel.