Software at CIS : CAWorks : Modules
Modules
Overview
Existing CAWorks command line programs. The modules are still being
tested. They will be used in the design of the next generation CAWorks
application.
Key:
<Required parameters>
[Optional parameters]
Converts 3D landmarks to 2D landmarks and 2D vice versa.
Input can be either CA .lmk format (landmarks) or .lmi format (index)
The user also has an option of outputing an index file (.lmi) if the input is a landmark file (.lmk).
Usage: 3D2Dlmk <3Dsurface.byu> <2Dsurface_er.p> <infile> <option 1> <option 2> <outfile>
where:
infile: 3D or 2D landmark filename (.lmi or .lmk, include extension)
option 1: if inputfile is a 3D landmark file then option 1 param is "3D".
if inputfile is a 2D landmark file then option 1 param is "2D".
if inputfile is a index file then option 1 param is "ind".
option 2: if Landmarks are to be saved as an index file the param is "1".
if Landmarks are not to be saved as an index file the param is "0" (Enter "0" if the input is a .lmi index file).
outfile: Output file name (do not include any extension).
SSD (Surface to Surface Distance Extraction)
SD (Surface Distance)If any of the points in surface 1 are greater than "max distance" to surface 2, they are removed from surface 1. Write surface 1 to the outputFile.
Usage: SSD <surface 1> <surface 2> <output file> <distance>
Given a byu and its associated img and segmentation, calculate euclidean norm distances from voxels in img to closest surface vertices. Outputs are 2 float imgsUsage: SD <Surface.byu> <Volume.img> <Segmentation.img> <whitematter label> <output_distance file> <output_Indices file>
Distance from voxels to surface vertices. Takes an sMap generated by SD, and
a 3 Peak segmentation file.
Outputs are 3 .txt files containing raw distance data of GM/WM/CSF distances.
Usage: ANTSY3P <Segmentation.img> <Smap.img> <RawDistance>
[Seg Tags: BG CSF GM WM]
where:
RawDistance: Output file name (e.g. 1234_L_ant)
Default Segmentation Tag Values:
BG = 0
CSF = 25
GREY = 75
WHITE = 125
Write the surface area and the volume of a surface to a file.
Reads in a Surface File, determines the Surface Area and the Volume, then
writes, or appends, these values to the specified output file.
Usage: SurfStats <Surface.byu> <Stats output>
[Output file options: 0 = overwrite, 1 = append (default = 0)]
[X pixel size (default = 1.0)]
[Y pixel size (default = 1.0)]
[Z pixel size (default = 1.0)]
where:
Stats output: Enter output file name (e.g. 1234_L_ant.txt)
Extract a surface from the intersection of a surface and an image and/or
segmentation. Reads in a Surface File, Image File, and Segmentation File.
Determines, based on Image and Segmentation Files, and user parameters,
which triangles should be discarded.
Usage: SurfaceExtractRegion <Surface.byu> <image_vol.img> <segmentation.img> <ExtractedSurface.byu>
[Min X (def = 0)] [Min Y (def = 0)] [Min Z (def = 0)]
Note - for Max values, -1 indicates use size of volume/segmentation
[Max X (def = -1)] [Max Y (def = -1)] [Max Z (def = -1)]
[keep where segmentation non-zero (0/1, default 1)]
[keep where image non-zero (0/1, default 1)]
[keep where seg value equals ... (0/1, default 0)] [seg value (default 0)]
[keep where seg value not equal ... (0/1, default 0)] [seg value (default 0)]
[keep where segmentation neighbors non-zero (0/1, default 0)]
[keep where normals face non-zero in segmentation (0/1, default 0)]
[keep where normals face value in segmentation (0/1, default 0)] [seg value (default 0)]
[length of normals (default 1.0)]
Takes an image and a threshold value and creates a triangulated isocontour
surface. Writes the resulting surface to the specified output file.
Usage: SurfaceGeneration <Image.img> <IsoSurface Threshold> <IsoSurface.byu>
[Optional Downsample Values: x y z (def = 1.0)]
Surface vs. Surface Statistics. Calculates statistics on the distance data
between two surfaces.
Usage: SvSStats <Surface1.byu> <Surface2.byu> <Raw Stats output file> <Text Stats output file>
[Surface1 dX dY dZ pixel sizes (default = 1.0)]
[Surface2 dX dY dZ pixel sizes (default = 1.0)]
Creates an image of voxels from an input image that
are no greater than the maximum distance from the input surface mask.
Usage: MinDistFromSurfaceMask <Image.img> <SurfaceMask.byu> <MaskedImage.img> <MaximumDistance>
[Optional WM Threshold: (def = -1.0, i.e., don't use)]
Creates an image of voxels from an input image that
are no greater than the maximum distance from the input surface mask.
Usage: MinDistFromSurfaceMask <Image.img> <SurfaceMask.byu> <MaskedImage.img> <MaximumDistance>
[Optional WM Threshold: (def = -1.0, i.e., don't use)]
Reclassifies a 5 compartment bayesian segmentation into a 3 compartment segmentation using optimal thresholds derived from N-P likelihood ratio testing
Usage: RESEG <Volume.img> <Segmentation.img> <stats> <PV1> <PV2> <Resegmentation.img>
[Seg Tag Values: BG CSF CSF/Grey Grey Grey/White White]
where:
PV1 - Partial Volume 1: Enter 0 to use default value 0.835
PV2 - Partial Volume 2: Enter 0 to use default value 0.08
Default Segmentation Tag Values:
BG = 0
CSF = 25
CSF/GREY = 50
GREY = 75
GREY/WHITE = 100
WHITE = 125
SubSD (Sub-surface Distance)
Given a parent surface and a child surface, calculate euclidean norm distances of voxels in img that are associated with child surface onlyCM (CurveMatch)Usage: SubSD <SurfaceP.byu> <SurfaceC.byu> <Segmentation.img> <output_distance file>
Takes coordinates of 2 curves as inputs and returns their corresponding matched coordinates.
Usage: CM <Curve1> <Curve2> <coord1> <coord2> <#samp> <#lmk> <polyorder> <nghbsize> <a_coeff> <c_coeff> <t_coeff> [L];
where:
coord1: Output file that saves Curve1 landmarks
coord2: Output file that saves Curve2 landmarks
#samp: Number of Samples (CA default = 100)
#lmk: Number of Landmarks (CA default = 12)
polyorder: Polynomial Order (CA default = 8)
nghbsize: Neighborhood Size (CA default = 10)
a_coeff: Arclength Co-efficient (CA default = 1)
c_coeff: Curvature Co-efficient (CA default = 1)
t_coeff: Torsion Co-efficient (CA default = 1)
L = 1 saves Landmarks in CAWorks .lmk format (default = 0)
poly_Redist
Usage: poly_Redist <Surface.byu> <NewSurface.byu> <Redistribute alpha> <Redistribute iterations>
where:
redistribute alpha (CA default = 0.005)
redistribute iterations (CA default = 50)
newlmk
Given a set of landmarks defined on the original surface, find corresponding set of landmarks on the downsampled surface and calculate euclidiean distance between original and downsampled
landmarks.
Usage: newlmk <lmkfile.lmk> <downsampsrf.byu> <outputlmk> <Option 1>
where:
Option 1 saves Euclidean distances
1 Saves it to [outputlmk.txt].
0 Does not save the Euclidean Distances.
Curvature
Curvature - Generate Surface Curvature
Calculate the curvature of a surface and write it to a file
Usage: Curvature <curvature type> <surface.byu> <output curvature file>
[Curvature Depth (default = 2.0)]
[Surface dX dY dZ pixel sizes (default = 1.0)]
where:
curvature type (def = 0):
0 = Maximum
1 = Minimum
2 = Mean
3 = Gaussian
CAWORKS COMMAND LINE PROGRAM WISHLIST
Below is a list of key programs I have identified as being used extensively by many CA users, that can benefit from a scriptable version and have an anxious audience awaiting the arrival of their command line counterparts:
- surface generation - generates triangulated mesh for a user defined threshold - (now implemented as SurfaceGeneration)
- surface extraction - keeps vertices of largest connected component and deletes all other nodes
- mean curvature calculation - calculates mean curvature of a surface - (now implemented as Curvature)
- surface extract using regions – given a parent surface, segmentation and a corresponding child image, keep vertices that satisfy a set of Boolean criteria (e.g. where segmentation and/or image is non-zero) - (now implemented as SurfaceExtractRegion)
- L1 distance computation - calculates L1 distance between manual and automated segmentation
- Surface/Contour distance - calculates distance between hand contours (manually generated) and polygonal mesh
Last Modified: Monday, 25th April, 2011 @ 11:19am
