VCFRAME | VisionX V4 |
vcframe - chest frame of reference transformation vector
SYNOPSISvcframe if=<ifile> ig=<ribfile> ih=<ribroots> [of=<ofile>] [-w] [-v]
DESCRIPTIONvcframe computes the scale-location vector that is needed to map an image or an annotation into its standard chest frame of reference form.
A standard frame of reference (SFOR) is defined, in a standard three-dimensional coordinate system, by a cubic frame whose corners are located in points where each coordinates has values -1 or +1 and whose main axes will be anatomically oriented going, with respect to the scanned subject, from right to left, from front to back, from toes to head. A rigid linear mapping between an image and an SFOR is uniquely defined by the scale-location vector (SLV) - a vector made of six elements <xoff,yoff,zoff,xs,ys,zs> storing the offsets and the scale factors needed for the transformation. To map the current image coordinate to a point in the standard frame, the image should first be translated by <xoff,yoff,zoff> (in the same unit as the if= image resolution) then scaled by <xs,ys,zs>. The image is assumed to be an axial view in standard VisionX coordinates where the x axis increases moving to the right, the y axis increases moving upward (in the image).
A chest frame of reference (CFOR) is an SFOR whose limits are extracted from a chest scan (typically a whole-lung CT scan) and determined by the chest anatomy as detailed further on. Concerning the y axis, the CFOR standardizes the subject orientations as if they were supine. Concerning the z axis, as the images in a CT scan are ordered to have increasing z coordinate values when moving downwards, the scan and its CFOR representation are set to have their z axes going in opposite directions.
The current implementation relies on the output of vsribs to determine the limits of the CFOR. In particular the vertical extent (i.e. along the z axis) is derived from the rib starting point image - the og= output of vsribs. On the x, y, and z axis the limits of CFOR are respectively as follows:
|
− |
the left extent of the segmented ribs (x=-1) |
|||
|
− |
the righ extent of the segmented ribs (x=+1) |
|||
|
− |
the anterior extent of the segmented ribs (y=-1) |
|||
|
− |
the posterior extent of the segmented ribs (y=+1) |
|||
|
− |
the lowest rib root (as determined by vsribs) (z=-1) |
|||
|
− |
the topmost rib root (as determined by vsribs) (z=+1) |
By default, a wish (TCL) command that sets the variable slv to the six comma-separated values is returned. When of= is specified, the program embeds the SLV values into the output image (as an element of generic floats). It is also possible to get the SLV using the -v flag.
CONSTRAINTif= must be the image supplied as input to vsribs. ig= and ih= are the byte images, computed by vsribs, of the segmented ribs and the rib roots — vsribs output files of= and og= respectively.
OPTIONS|
if= |
input whole lung CT scan (as supplied to vsribs) | ||
|
ig= |
input segmented ribs (as provided by vsribs) | ||
|
ih= |
input segmented rib roots (as provided by vsribs) | ||
|
of= |
output image with SLV embedded (optional) | ||
|
-w |
default wish output (defines the slv variable containing the six comma-separated SLV values) | ||
|
-v |
verbose mode |
vsribs, vchmap, vbmkmap
AUTHORJ. Lee, AM Biancardi