X ray - Perform simple orthographic volume visualization

Name

	X ray - Perform simple orthographic volume visualization

Author

INFOMATION
	Larry Gelberg
	larryg@stardent.com
	Advanced Visual Systems Inc. (AVS Inc.)
	6 N.E.T.C, 521 Virginia Road, Concord, MA 01742 

Summary

	Name	X ray
	Type	Filter
	Inputs	input - 	field 3D 3-space 1-vector uniform byte REQUIRED
	Outputs	output - 	field 2D 2-space 1-vector uniform integer
	Parameters
	Name		Type	Default	Choices
	direction	choice	K	I,J,K	
	operation	choice	mean	sum,mean,median,min,max

Description

	"X ray" does simple, orthogonal volume visualization on 3D
uniform byte fields. It outputs a 2D integer field which can be
colorized and displayed as an image. Essentially, for a given
direction, the module looks at the row of voxels "behind" each screen
pixel and, depending of the selected operation, creates a new pixel
based on those voxels.
   The processing options are:
   SUM : each screen pixel is the sum of the stack of voxels
   MEAN : each screen pixel is the sum of the stack of voxels divided
	by the number of voxels in each stack 
   MEDIAN : the stack of voxels is sorted by value and the screen
	pixel gets the center value in the sorted stack
   MIN : the screen pixel gets the smallest value in the stack
   MAX : the screen pixel gets the largest value in the stack
  	"X ray" is a pretty fast volume visualization technique and
you should be able to easily extend it to cover other cases.

Inputs

	input - field 3D 3-space 1-vector uniform byte
This is essentially the ".dat" format.

Parameters

direction (radio_buttons)
	The choices are I, J, and K. The default is K. If you have
chosen I, you look down the Y axis into the XZ plane. If you have
chosen J, you look down the X axis into the YZ plane. If you have
chosen K, you look down the Z axis into the XY plane. 
operation (radio_buttons)
	The choices are sum, mean, median, min, and max. The default
is mean. MEDIAN is very slow to compute and produces results of
limited utility. SUM and MEDIAN produce the same visual results if you
normalize the colormap to the data, but MEDIAN takes a little longer
to compute. MIN is not very useful for medical data.
Your best bets are MEAN and MAX for most operations.

Outputs

	output - field 2D 2-space 1-vector uniform integer
"X ray" outputs an integer field since the sum operation can produce
values greater than 255. This must be colorized in order to be
displayed.

Example networks

Here is the typical X ray network:
	GENERATE COLORMAP		READ FIELD
		|			    |
		|			  X RAY
		|_____________	  __________|	
			     |    |         |
			  COLOR RANGE       |
                               |_________   |
					|   |
				      COLORIZER
					  |
			            DISPLAY IMAGE

Related modules

	TRACER, ORTHOSLICER

See also