Difference between revisions of "2015 Winter Project Week:SlicerMicroMacroScale"

From NAMIC Wiki
Jump to: navigation, search
(Created page with '__NOTOC__ <gallery> Image:PW-2015SLC.png|Projects List </gallery> ==Key Investigators== * Nicole Aucoin (BWH) * Bradley Lowenkamp (Medical…')
 
 
(11 intermediate revisions by one other user not shown)
Line 2: Line 2:
 
<gallery>
 
<gallery>
 
Image:PW-2015SLC.png|[[2015_Winter_Project_Week#Projects|Projects List]]
 
Image:PW-2015SLC.png|[[2015_Winter_Project_Week#Projects|Projects List]]
 +
Image:CtkVTKSeedsScaling.jpg|CTK example showing seed widget scaling between slice view and three dimensional view.
 +
Image:SlicerSeedScaling.jpg|Slicer example with 2d fiducial manager scaling set to 1.0 instead of 0.003
 
</gallery>
 
</gallery>
  
Line 7: Line 9:
  
 
* Nicole Aucoin (BWH)
 
* Nicole Aucoin (BWH)
* Bradley Lowenkamp (Medical Science Computing)
+
* Bradley Lowekamp (Medical Science Computing)
 +
* Jim Miller (GE)
  
 
==Project Description==
 
==Project Description==
Line 21: Line 24:
 
** [http://na-mic.org/Mantis/view.php?id=3547 Fiducials disappear with large image spacing]
 
** [http://na-mic.org/Mantis/view.php?id=3547 Fiducials disappear with large image spacing]
 
** [http://na-mic.org/Mantis/view.php?id=3657 Setting the unit to microns doesn't help]
 
** [http://na-mic.org/Mantis/view.php?id=3657 Setting the unit to microns doesn't help]
 +
** [http://na-mic.org/Mantis/view.php?id=3713 landmark size in 3D view not consistant with 2D views]
 
* Sample data sets
 
* Sample data sets
 
** [https://www.dropbox.com/s/a8ie3eww7ssis2u/E12.5-3747-r7__rec.7z Small]  - link out of date, need new sample from Murat
 
** [https://www.dropbox.com/s/a8ie3eww7ssis2u/E12.5-3747-r7__rec.7z Small]  - link out of date, need new sample from Murat
Line 28: Line 32:
 
<div style="width: 27%; float: left; padding-right: 3%;">
 
<div style="width: 27%; float: left; padding-right: 3%;">
 
<h3>Progress</h3>
 
<h3>Progress</h3>
*
+
* Investigating the base scaling assumptions for vtkSeedWidgets in 3d ctkVTKRenderView vs 2d ctkVTKSliceView
 +
** writing a pure CTK test case - it shows that in the default set up for the two kinds of views and mirrored seed widgets the seeds are the same size (large)
 +
** in Slicer, the camera in the 2d view is looking at the 2d scene from a camera distance of 1.0 while the default 3D distance is 600.0
 +
** in the pure CTK test case, resetting the 3D camera to a position (0,0,600) the same issues can be seen as are visible in Slicer (tiny seed in 3D, large in 2D)
 +
** talking with Jim:
 +
*** when background volume changes, reset the camera position for each slice view to take into account the bounding box of the data
 +
*** use a distance calculation formula that takes the viewing angle into account (default 30):  (max_data_dim/2.0) / sin(view_angle / 2.0)
 +
** once have a new setting:
 +
*** reset the scale factor 2d on the 2d fid disp manager to 1.0
 +
*** reset the Z distance of the camera position in the slice view
 +
**** make sure set the camera on the correct renderer
 +
***** there is an overlay renderer on the slice viewers so using the first renderer from the collection in my ctk example didn't work
 +
***** get the renderer from the light box manager at all times, a light box index of 0 is always fine
 +
***** turns out that when creating a new seed, you need to explictly reset the renderer on the new handle
 +
***** lost ability to manipulate seeds, need more debugging
 +
*** trigger rerendering of the seeds (in tests, we needed to remove all and readd them to see the new size)
 +
*** double check placing rulers and ROIs in 2d
 +
*** double check display to world coordinate calculation in the 2d disp managers
 
</div>
 
</div>
 
</div>
 
</div>

Latest revision as of 15:08, 12 January 2015

Home < 2015 Winter Project Week:SlicerMicroMacroScale

Key Investigators

  • Nicole Aucoin (BWH)
  • Bradley Lowekamp (Medical Science Computing)
  • Jim Miller (GE)

Project Description

Objective

  • Fix bugs exposed by using small and large scale images

Approach, Plan

Progress

  • Investigating the base scaling assumptions for vtkSeedWidgets in 3d ctkVTKRenderView vs 2d ctkVTKSliceView
    • writing a pure CTK test case - it shows that in the default set up for the two kinds of views and mirrored seed widgets the seeds are the same size (large)
    • in Slicer, the camera in the 2d view is looking at the 2d scene from a camera distance of 1.0 while the default 3D distance is 600.0
    • in the pure CTK test case, resetting the 3D camera to a position (0,0,600) the same issues can be seen as are visible in Slicer (tiny seed in 3D, large in 2D)
    • talking with Jim:
      • when background volume changes, reset the camera position for each slice view to take into account the bounding box of the data
      • use a distance calculation formula that takes the viewing angle into account (default 30): (max_data_dim/2.0) / sin(view_angle / 2.0)
    • once have a new setting:
      • reset the scale factor 2d on the 2d fid disp manager to 1.0
      • reset the Z distance of the camera position in the slice view
        • make sure set the camera on the correct renderer
          • there is an overlay renderer on the slice viewers so using the first renderer from the collection in my ctk example didn't work
          • get the renderer from the light box manager at all times, a light box index of 0 is always fine
          • turns out that when creating a new seed, you need to explictly reset the renderer on the new handle
          • lost ability to manipulate seeds, need more debugging
      • trigger rerendering of the seeds (in tests, we needed to remove all and readd them to see the new size)
      • double check placing rulers and ROIs in 2d
      • double check display to world coordinate calculation in the 2d disp managers