Center for Imaging Science : About | Research | Publications | Education | Activities | Downloads | Visiting
BIRN Header home

Software @ CIS : lddmm-volume : Birn Portal

about | tutorial | user validation | manual | birn portal | namespace | faq | credits | changelog | feedback

New Users

If this is the first time you have visited this web page, the Introduction section below would be a good place to start.
Content

Example DATA

Introduction

The lddmm BIRN Portal is being developed to enhance the lddmm processing and analysis of image data. This tutorial is aimed at persons wishing to analyze shape data using lddmm.

Under the auspices of the Brain Morphometry Biomedical Informatics Research Network (mBIRN), a processing pipeline (SASHA - Semi-Automated Shape Analysis) is being developed to enable seamless processing of brain morphometry data for subcortical structures through the integration of multiple site applications The components of SASHA are:

LDDMM - software that allows for comparison and quantization of morphometric changes in shapes.

Free Surfer - set of semi-automated tools for reconstruction of the brain's cortical surface and overlay of functional data onto the reconstructed surface.

3D Slicer - open-source application that assists with the visualization, registration, segmentation, and quantification of medical image data.

This tutorial covers use of lddmm on the BIRN Portal and viewing the results. At the end of the tutorial a user will be able to run lddmm on the BIRN Portal and visualize the results. In the tutorial the user will access the LDDMM BIRN portal, select two hippocampus volumes to compare, Run lddmm and visualize the output. For further information on lddmm refer to Clinical Background.

This can require large computational resources, depending on the images used for comparison. The study referenced in Clinical Background was performed using the TeraGrid.

Click on any of the images in this tutorial to display a larger version of the image.

Login

Login into the LDDMM BIRN Portal.

Select Data Management

The image below shows the data management selection. The below image views the home directory where sample data is located.

Note

  • LDDMM processing using the sample hippocampus data will take approximately three hours.
  • LDDMM processing using the sample CVRG Cardiac data will take approximately 5 minutes.

Select Data

Note that the data directory contains a collection of images. There is one Atlas (Template) and several Patient (Target) data sets. The BIRN portal access the data files throught the (Storage Resource Broker).

Select Analysis Tools

Select the 'rocketship' icon in order to select analysis tool options.

Select LDDMM Tool

Select the LDDMM tool. Notice that the right side of the image allows you to select an image.

Select Template

Select the 'Template' for processing. This lddmm processing will compare this template image to all other images in the give working directory. It is important to have all images you wish to compare to a Template in a single directory.

Select Launch

After selecting the the Template (Atlas) and the Launch button, a dialog box will appear to ensure that you wish to proceed. The dialog box will provide information on the Template (Atlas) and the dimension of the image.

Job Status

This window provides status information regarding all jobs being processed.

Detail Job Status

This option provide detailed information relating to a job submission. It provide the detailed command line, user Proxy information and status of all the comparisons.

Email Notification

An email notification is provide when the job is launched. This will allow you to verify the server deamon has accepted your request for processing.

View Jobs Running

This shows the most current job has 7 jobs running.

View Jobs Completing

This shows the most current job has 3 jobs running and 4 jobs are complete.

View Email Results

When a job is complete, an email is sent which provides the finalMetricDistance information.

View Result Directory

View Target Directory

The input and output data for LDDMM on the BIRN Portal is accessed using the (Storage Resource Broker). Two interfaces exist to view the SRB data, InQ and SCommands. InQ is a Windows based program and is available from

InQ

InQ is used to explore data located on the SRB. You will need to configure the InQ in order to download the resultant data in order to visualize the lddmm data.

InQ Directory

The below screen shot diplays the metadata and directory structure of ldmm. The three resulant directories are vtk, txt, and analyze. The vtk directory contains the velocity data and the txt provides logging information and the metric distance between the two images.

Get Data through Scommands

The following is an example of retrieving data through the SRB Scommands. A local copy of the vtk data is needed for visualization. Note the metric distance in this comparison is 4.3686689. 
[anthony@jhu-gpop anthony]$ Sinit
[anthony@jhu-gpop anthony]$ Scd /home/akolasny.jhu-cis/renewal/XRG_lh.hippocampus_96
[anthony@jhu-gpop anthony]$ Sls
/home/akolasny.jhu-cis/renewal/XRG_lh.hippocampus_96:
  C-/home/akolasny.jhu-cis/renewal/XRG_lh.hippocampus_96/analyze
  C-/home/akolasny.jhu-cis/renewal/XRG_lh.hippocampus_96/txt
  C-/home/akolasny.jhu-cis/renewal/XRG_lh.hippocampus_96/vtk
[anthony@jhu-gpop anthony]$ Scat txt/finalMetricDistance.txt
4.3686689
[anthony@jhu-gpop anthony]$ Scd vtk
[anthony@jhu-gpop anthony]$ Sls
/home/akolasny.jhu-cis/renewal/XRG_lh.hippocampus_96/vtk:
  C-/home/akolasny.jhu-cis/renewal/XRG_lh.hippocampus_96/vtk/deformation_maps
  C-/home/akolasny.jhu-cis/renewal/XRG_lh.hippocampus_96/vtk/deformed_target
  C-/home/akolasny.jhu-cis/renewal/XRG_lh.hippocampus_96/vtk/deformed_template
  C-/home/akolasny.jhu-cis/renewal/XRG_lh.hippocampus_96/vtk/inverse_deformation_maps
  C-/home/akolasny.jhu-cis/renewal/XRG_lh.hippocampus_96/vtk/other
  C-/home/akolasny.jhu-cis/renewal/XRG_lh.hippocampus_96/vtk/velocity_fields
[anthony@jhu-gpop anthony]$ Scd /home/akolasny.jhu-cis/renewal
[anthony@jhu-gpop anthony]$ Sget -b XRG_lh.hippocampus_96 .

Visualize using MayaVi

MayaVi provides a helpful gui interface to visually explore the velocity vectors. Loading two hippocampus data sets and the velocity vectors, we see the areas of change between the structures.

Using MayaVi

When MayaVi is started the following window is displayed. Menu items in MayaVi can be selected using a mouse or keystrokes. In the following sections keystrokes are listed in parentheses.

To import the first data set select the File menu (Alt + F). The file menu bar is then displayed.

Select Open (O) on the file menu. A second menu listing file open options is displayed.

Select Vtk File (V) on the this menu. A file selection dialog is displayed.

Navigate to the appropriate directory (deformed_target) and select the desired *.vtk file (MoviedefPatient19.vtk). The data visualizer screen is displayed.

Select Visualize (Alt + V). The visualize menu is displayed.

Select Modules (M). The modules menu is displayed.

Select Isosurface (I). The image is displayed in the Visualization portion of the MayaVi window.

Perform the following actions on the Configure IsoSurface module dialog:

1) check the Compute PolyDataNormals checkbox.

2) Set Opacity to 0.55 using the Set Opacity scroll bar.

3) Close the Configure IsoSurface module dialog.

The first data set is now imported.

Import the second data set using the same procedure as the first data set:

1) File (Alt + F)

2) Open (O)

3) Vtk File (V)

4) Navigate to the appropriate directory (deformed_template) and select the desired *.vtk file (MoviedefAtlas0.vtk)

The first image is still displayed.

Navigate to the Configure IsoSurface module dialog as before:

1) Visualize (Alt + V)

2) Modules (M)

3) Isosurface (I)

Perform the following actions on the Configure IsoSurface module dialog:

1) check the Compute PolyDataNormals checkbox.

2) Deselect the Scalar Coloring checkbox.

3) Select Change Object Color. The Change Object Color dialog is displayed. The Change Object Color dialog differs from Windows to Unix. The Windows dialog is discussed first, followed by unix dialog.

1) Select the desired color using mouse.

1) The desired color parameters are displayed in the edit boxes on the right hand side of the dialog. The color selection also can be input using these edit boxes. Select OK to close the Change objects color dialog.

1) Unix users can enter the desired color as follows. Place cursor in Selection box, and delete value #ffffff. Enter value #007ec0 (hit Return Key after entering this value, before selecting OK button). Select the OK button to apply the color change.

1) The two images can now be clearly distinguished. Set Opacity to 0.55 using the scroll bar and close the Configure IsoSurface Module dialog.

The second data set is now imported.

Import the third data set using the same procedure as the first data set:

1) File (Alt + F)

2) Open (O)

3) Vtk File (V)

4) Navigate to the appropriate directory (velocity_fields) and select the desired *.vtk file (MovieVelocity19.vtk)

Select Visualize (V). The visualize menu bar is displayed.

Select Filters (F). The filter menu is displayed. Select ExtractVectorNorm (V).

The Configure Vector dialog is displayed. Select OK to close the Configure Vector dialog.

Select Filters (F). The filter menu is displayed. Select Threshold (H).

The Configure Threshold dialog is displayed. Enter 2.0 in Minimum threshold edit box. Hit [Return] before selecting Close button. Select Close button on Configure Threshold dialog.

The first two data sets are displayed.

Select Visualize (V) Modules (M) VelocityVector (Y).

The two hippocampus data sets and velocity vectors are now displayed.

Visualize using ParaView

ParaView provides a helpful gui interface to visually explore the velocity vectors. Loading two hippocampus data sets and the velocity vectors, we see the areas of change between the structures.

Using ParaView

When ParaView is started the following window is displayed. Menu items in ParaView can be selected using a mouse or keystrokes. In the following sections keystrokes are listed in parentheses.

To import the first data set select the File menu (Alt + F). The file menu bar is then displayed.

Select Open Data (O) on the file menu. A file selection dialog is displayed.

Navigate to the appropriate directory (deformed_template) and select the desired *.vtk file (MoviedefAtlas0.vtk). The data visualizer screen is displayed.

Move cursor to the "Generate Isolines or Isosurfaces" icon. This is the sixth icon from the left on the toolbar.

Select the icon using the left mouse button. The modules menu is displayed.

Select the "Accept" button. The Isosurface is displayed in the 3D View window.

Select the Display tab.

Locate the "Actor Color" dialog (in the Color group).

Select a color from palette (or enter #fbe966 and then select "Enter") and press the "OK Button".

Set "Opacity" to 0.5 in the lowermost control.

Import the second data set using the same procedure as the first data set:

1) File (Alt + F)

2) Open Data (O)

3) Navigate to the appropriate directory (other) and select the desired *.vtk file (Patient.vtk) and select the IsoSurfaces icon.

To zoom in press "Ctrl" and the left mouse button while the cursor is in the viewing area.

The first image is still displayed.

As before, select the "Display" tab, change the color of the second isosurface (#4c4959 in this example), and set opacity to 0.2. Change Display Style from Surface to Wireframe using "Representation" Button.

Import the third data set using the same procedure as the first data set:

1) File (Alt + F)

2) Open Data (O)

3) Navigate to the appropriate directory (velocity_fields) and select the desired *.vtk file (MovieVelocity29.vtk). For large data set, this may take some time (tens of seconds).

Select "Generate a Glyph" icon. This is the tenth icon from the left on the toolbar.

Select the "Accept" button. The template and target isosurfaces are displayed, as well as velocity vectors.

To display glyphs in color, select "Display" tab, Color by: "Poing GlyphVector (3)" button.

The number of glyphs can be modified using the "Parameters" tab, Max. Number of Glyphs edit box. The user can rotate the image using the mouse with left button depressed.

Legacy Portal: Documentation regarding the older version of the BIRN portal.

Last Modified: Wednesday, 05-Sep-2007 15:18:05 EDT

NIH                    NCRR                     DHHS                      TeraGrid

 
 




301 Clark Hall
3400 N. Charles Street
Baltimore, MD 21218
Office: (410) 516-3826
Fax: (410)516-4594
webmaster@cis.jhu.edu