Reformat

MLModule

genre

Resample

author

MeVis Medical Solutions AG

package

MeVisLab/Standard

dll

MLResample1

definition

MLResample1.def

see also

SoView2DOverlay, MPR, MPRPath

keywords

interpolation, coordinate, system, matrix, transform, mapping

Purpose

The module Reformat reformats an image to a reference image and produces reformatted overlays.

Usage

Using Reformat for reformatting an image to a reference image

This module is used to reformat the dataset given as input0 to the local coordinate system of a reference image input1. The output image output0 has the same image properties as input0, but contains the reformatted image data of image0. An additional matrix is used to rotate/scale/translate the input image.

Using Reformat for overlay images

This module can also generate overlay images for any given view of a SoView2D / SoOrthoView2D. This works by connecting output1 to an SoView2DOverlay module and connecting the SoView2DOverlay back to the overlay input.

By doing this, the SoView2DOverlay tells the Reformat about the redraw and output1 is reformatted to the image currently drawn in SoView2D / SoOrthoView2D.

../../../Modules/ML/MLResample1/mhelp/Images/reformat1.jpg

Interaction

Reformat also has an interactive mode for manipulation of the overlay image, which can be used for interactive adjustment of the additional matrix.

For using this, you have to select a mouse button in the SoView2DOverlay panel. The image is translated by pressing this mouse button and rotated around the z-axis of the current view while holding the SHIFT key.

On the second tab of the module’s window, the mouse buttons, keys, and modifier keys can be defined for an interactive translation and/or rotation.

Tips

This module is optimized using 16.16 fixed-point arithmetic for data types with 16- and 8-bit!

It also works for other data types but uses a much slower double precision method, which currently only supports trilinear interpolation.

Windows

Default Panel

../../../Modules/ML/MLResample1/mhelp/Images/Screenshots/Reformat._default.png

Input Fields

input0

name: input0, type: Image

Input image to be reformatted.

input1

name: input1, type: Image

Reference image.

inOverlay

name: inOverlay, type: SoNode, deprecated name: overlay

Input for the output of a SoView2DOverlay module for interactive use.

Output Fields

output0

name: output0, type: Image

Reformatted image.

output1

name: output1, type: Image

Special output for SoView2DOverlay.

Parameter Fields

Field Index

Additional Matrix: Matrix

Interpolation: Enum

Apply: Trigger

Memory Access: Enum

Auto center: Bool

Reset: Trigger

Border Handling: Enum

Rotation: Rotation

Center: Vector3

Scale: Vector3

Compose matrix: Bool

Translation: Vector3

Fill Value: Double

Update Mode: Enum

Interactive: Bool

useInventorComposition: Bool

Visible Fields

Additional Matrix

name: matrix, type: Matrix, default: 1 0 0 0, 0 1 0 0, 0 0 1 0, 0 0 0 1

Sets a matrix representation of transformation components.

This parameter can also be modified manually or via field connection.

Compose matrix

name: composeMatrix, type: Bool, default: TRUE

If checked, the (interactive or manual) manipulation of matrix components is activated.

For the interactive mode, the option Interactive needs to be checked.

Scale

name: scale, type: Vector3, default: 1 1 1

Sets a scale factor vector that is applied to the input image.

Rotation

name: rotation, type: Rotation, default: 0 0 1 0

Sets a rotation that is applied to the input image.

Translation

name: translation, type: Vector3, default: 0 0 0

Sets a translation that is applied to the input image.

Center

name: center, type: Vector3, default: 0 0 0

Sets the center for the Rotation.

Fill Value

name: fillValue, type: Double, default: 0, deprecated name: fillvalue

Sets the value used for regions outside of the input image. The value is automatically clamped to the value range of the voxel data type of the first input image.

Interactive

name: interactive, type: Bool, default: TRUE

If checked, the module accepts interactive manipulation (with a SoView2DOverlay module).

Auto center

name: autoCenter, type: Bool, default: TRUE

If checked, the center of rotation is always the center of the image.

Interpolation

name: interpolation, type: Enum, default: Trilinear, deprecated name: Mode

Defines the interpolation for resampling.

Values:

Title

Name

Deprecated Name

Description

​Nearest Neighbor

​NearestNeighbor

​Nearest Neighbor

​No interpolation, picks the nearest input image voxel. If the decision is ambiguous (the center of a target voxel lies exactly on the boundary between two source voxels), the value will be taken from the source voxel with the higher coordinate value (i.e., the right one in 1D).

Fastest mode.

​Trilinear

​Trilinear

​Trilinear(skip border),TrilinearSkipBorder

​Trilinear filtering. The 8 voxel neighborhood is used for linear interpolation.

Slowest mode.

Border Handling

name: borderHandling, type: Enum, default: UseFillValue

Defines how borders are handled in trilinear filtering.

Values:

Title

Name

Description

​Clamp To Edge

​ClampToEdge

​Clamps the interpolation to the edge of the input volume and uses the fill value only for values beyond the edge.

../../../Modules/ML/MLResample1/mhelp/Images/TrilinearClampToEdge.png

​Use Fill Value

​UseFillValue

​Uses the fill value for values on the border and outside of the volume.

../../../Modules/ML/MLResample1/mhelp/Images/TrilinearUseFillValue.png

Memory Access

name: memoryAccess, type: Enum, default: Global

Defines how the module accesses the input image.

Values:

Title

Name

Description

​Global

​Global

​Fastest, but highest memory consumption and initialization time. The input image is accessed as a global image (via the ML memory image). This consumes a lot of memory for large datasets.

If not enough memory is available, the module falls back to the Virtual Volume memory access.

​Virtual Volume

​VirtualVolume

​Slowest, but works on huge datasets. All pages of the input image are requested via the virtual volume mechanism, thus only small portions of the input image are needed at the same time.

Reset

name: resetMatrix, type: Trigger

When pressed, the fields Scale, Translation, Center, and Rotation are reset to their default values.

Update Mode

name: updateMode, type: Enum, default: AutoUpdate

Defines the update mode of this module.

Values:

Title

Name

Description

​Update

​AutoUpdate

​On any input or parameter change, the output image is updated.

​Clear

​AutoClear

​On any input or parameter change, the output image is invalidated. Consecutive changed do not fire any notification.

Apply

name: apply, type: Trigger

When pressed, the output image is updated according to the current input images and parameter settings.

Hidden Fields

useInventorComposition

name: useInventorComposition, type: Bool, default: FALSE