Loading 3D Data into Analyze

The 3D rendering program will work better on datasets that have been preprocessed to be properly registered and segmented. Registration and segmentation removes obscuring structures and unnecessary artifacts that tend to clutter the final surface. The cleaned up data can then be saved as an object map that will be used by the 3D surface tiler.

If multiple structures are to be rendered, first isolate each structure using the segmentation tools, and then render each surface individually. The resulting surfaces can then be imported into Softimage for later compositing and rendering.

The surface generation algorithm will work on non-segmented data, and it is that type of data used in the examples found in this tutorial.

Some of the surface visualization algorithms work best on larger datasets. If the after following the steps in this tutorial, the visualization algorithm produces poor results, try scaling the data. But understand that the interpolated data cannot be considered "pure", and should not be used for mensuration or analysis.

Do it yourself...
1.	For additional help, click on the thumbnail images on these pages to view a full-sized version of each image. These images are direct screen captures of the AnalyzeAVW interface as it appears on the ITG UNIX workstations. Use them as a reference in case of confusion.
2.	To use the dataset in this example, download the file glurigor.avw from the ITG web server. The dataset is approximately 750K, so be prepared for a long download time over very slow connections. glurigor.avw has been made available courtesy Dr. Bridget Carragher (Carragher 15206).
3.	In the VMIL, pick a machine on which to run Analyze. Available systems include one Windows NT system - hatoum - and three UNIX workstations - zeus, delphi, and claros. delphi is a suitable choice and a powerful machine.
4.	Launch Analyze by typing Analyze at the UNIX command prompt (or by selecting AnalyzeAVW from the Start menu if using Windows NT instead of UNIX).
5.	From the main workspace menubar, use File:Load As... to load the dataset.
6.	Press the File button in the upper left-hand corner of the dialog box, and locate the locally saved file, selecting it with clicks of the left mouse button. Or, type the full path name to the file in the text field.
7.	Toggle the Resize feature on, and multiply each of the axis-aligned dimensions by three (X: 219, Y: 303, Z: 303).
8.	For best results, turn on the Interpolate feature.
9.	The Resize on/off switch should change to a green color indicating that a transformation operation will take place during the load.
10.	Press the Load button when done. Listing the contents of the directories and loading may take a few seconds. While loading a progress bar should pop up indicating the percentage of data read.
11.	Ensure the dataset is selected by clicking on it once with the left mouse button. A red outline appears around the image icon to indicate the current selection(s).

Page 3 of 12