FilterConnectedComponents

MLModule

author	MeVis Medical Solutions AG
package	MeVisLab/Standard
dll	MLCluster
definition	MLCluster.def
see also	ComputeConnectedComponents, ConnectedComponentsToImage, ConnectedComponentsInfo
keywords	region, segmentation, cluster, neighborhood, select

Purpose

The module FilterConnectedComponents filters and selects certain clusters that were computed by a connected component analysis.

Details

The input clusters are first filtered by their size. The filtering is optional. By default, all clusters are used for selection.

Then, a selection can be performed on the resulting cluster subset. All background voxels are stored in one cluster which cannot be selected.

The input cluster structure consists of voxel-based cluster information and a selection part. The output cluster structure shares the voxel-based cluster information of the input structure but yields its own selection information. So the output structure is a copy of the input structure but the memory-intensive clusters themselves are just shared.

Windows

Default Panel

Input Fields

inClusters

name: inClusters, type: SelectedClusters(MLBase)

For accessing this object via scripting see the Scripting Reference: MLSelectedClustersWrapper.

inMarkerList

name: inMarkerList, type: XMarkerList(MLBase)

For accessing this object via scripting see the Scripting Reference: MLXMarkerListWrapper.

Output Fields

outClusters

name: outClusters, type: SelectedClusters(MLBase)

For accessing this object via scripting see the Scripting Reference: MLSelectedClustersWrapper.

Parameter Fields

Field Index

Clear: Trigger

Cluster Size Interpretation: Enum

Invert selection: Bool

Max Cluster Size: Double

Min Cluster Size: Double

Selection Mode: Enum

Update: Trigger

Update Mode: Enum

Visible Fields

Update Mode

name: updateMode, type: Enum, default: AutoClear

Defines the update behavior of this module.

Values:

Title	Name	Description
Clear	AutoClear	On any input or parameter field change, the module clears its output.
Update	AutoUpdate	On any input or parameter field, the module filters/selects the output cluster structure anew.

Update

name: update, type: Trigger

If pressed, the module computes anew.

Clear

name: clear, type: Trigger

If pressed, the module’s output is cleared.

Min Cluster Size

name: minClusterSize, type: Double, default: 0, minimum: 0

Sets a minimum cluster size for filtering. See Cluster Size Interpretation.

A value of 0 mean no filtering by a minimum size.

Max Cluster Size

name: maxClusterSize, type: Double, default: -1, minimum: -1

Sets a maximum cluster size for filtering. See Cluster Size Interpretation.

A value of -1 means no filtering by a maximum size.

Cluster Size Interpretation

name: clusterSizeInterpretation, type: Enum, default: Voxels

Defines how the minimum and the maximum cluster size for filtering should be interpreted.

Values:

Title	Name	Description
Voxels	Voxels	The min and max sizes are interpreted as number of voxels.
ml	ml	The min and max sizes are interpreted as milliliters. The cluster structure knows how many milliliters a voxel has.

Invert selection

name: shouldInvertSelection, type: Bool, default: FALSE

If checked, the selection is inverted. Note that the background is never selected.

Selection Mode

name: selectionMode, type: Enum, default: AllFiltered

Defines the selection mode.

Values:

Title	Name	Description
All Filtered	AllFiltered	Sets all filtered clusters as selected.
Selected By Coordinates	SelectedByCoordinates	Sets those clusters selected where at least one position of the module’s inMarkerList lies in a voxel. The markers are given in world position, the module converts those positions to voxel positions. Note that every cluster will only be selected once, even if more than one marker lies within the cluster. Also note that the background is never selected.
Largest	Largest	Selects the largest cluster. If there are multiple largest clusters with the same size, the first of these clusters is selected.
Largest On Slice	LargestOnSlice	Selects the largest cluster on each slice. Note that this mode is only available if the cluster algorithm was using a 2D neighborhood. If this option is chosen but the cluster algorithm was using a 3D neighborhood, an error message is printed to the console and the module does not select any cluster.