Diffusion Tensor Image(DTI) Volumes
Scuba2 can view DTI volume slices with varying levels of detail and opacities. Each voxel contains a tensor pointing in the direction of the major eigenvector, also encoded by the color:
Brightness is weighted by fractional anisotropy(FA), where brighter tensors indicate high FA and dimmer tensors indicate low FA.
After launching scuba2, load a DTI volume by:
Go to File->Load DTI... This will open a file browsing window.
- Navigate to the directory containing your tensor data. Make sure that the following volumes are in the same directory: fa, eigenvals, eigenvec1, eigenvec2, eigenvec3.
Select the fa volume and click Open.
This will load your DTI volume as slices of tensors.
In the Display Layer you have two modes of interaction, 2D and 3D. Both modes display slices of the diffusion tensor volume and enable you to move through the X, Y, and Z slices. You can switch between 2D and 3D mode by selecting either the 2D or 3D radio buttons in the View panel. In the 2D and 3D mode you can configure the following:
Opacity: Changes the transparency of the tensor visualization.
Interpolation: Changes the method of smoothing and resolving the new data points (tensors). The three options are:
Linear: Smooths tensors linearly.
Cubic: Smooths the tensors to the second derivative.
Nearest Neighbor: No smoothing, uses the closest data point to the voxel as the tensor to display.
Draw Edges: Turns on and off a dark bounding box around each tensor.
Scaling: Scales the tensor on FA or uniformally scale them. The two options are:
Uniform: Scales tensors so that they're approximately the same size.
FA: Scales tensors so that voxels with high FA have large tensors and voxels with low FA have small tensors.
Detail: Varies the level of detail of the displayed slices. At levels other than Normal tensors will be averaged together so that fewer tensors are rendered, but the visualization is more interactive. The three settings are:
Least: Least level of detail and highly interactive.
Less: More than Least and less detail Normal.
Normal: Most level of detail and least interactive.
The following keys can be used:
R -- Reset: Resets the camera to the original position. Use this if you become lost in your visualization or want to zoom out if you're too far in.
F -- Focal Point: Sets the focal point of the camera to that of where the mouse is. If this is over an object, then the point is set to the surface of that object. This will enable you to zoom into that point and rotate about it.
In 2D mode (default), in the Display Mode Controls, you can click on the X, Y, or Z buttons to switch view planes. You can move through the slices by sliding the scroll bar or entering in the slice number. In 2D mode, the mouse can be used for navigating the data in the following manner:
Left Button: Pan slice.
Middle Button: Moves through slices.
Right Button: Zoom into slice.
In 3D mode, you can modify all three view planes without switching between them using the buttons. In the Layer panel, you have the following additional options:
Show * RAS Tensors (where * is X, Y, or Z): This makes the view planes visible or invisible.
Show all tensors: (currently disabled) Shows all tensors in the volume.
In 3D mode, the mouse can be used for navigating the data in the following manner:
Left Button: Pan visualization.
Middle Button: Rotate visualization.
Right Button: Zoom visualization.
- Make sure that the right volumes, named correctly, are in the same directory of the fa volume that you load.
Use Nearest Neighbor Interpolation for tensor truest to the data--nothing will be interpolated.
When using Normal detail, turn off slices that you're not interested in for a more interactive program.