Projects:RegistrationLibrary:RegLib C21

From NAMIC Wiki
Revision as of 15:21, 6 September 2011 by Meier (talk | contribs) (→‎Download)
Jump to: navigation, search
Home < Projects:RegistrationLibrary:RegLib C21

Back to ARRA main page
Back to Registration main page
Back to Registration Use-case Inventory

v3.6.3 This case is complete and up to date for version 3.6.3 Slicer Registration Library Case #21: Intra-subject knee MRI

Input

this is the fixed PET/CT image. All images are aligned into this space lleft this is the moving image. The transform is calculated by matching this to the reference image
fixed image/target moving image

Modules

Objective / Background

atlas building, priors, cartilage segmentation

Keywords

knee, MRI

Input Data

  • reference/fixed : subject 58 MRI: 0.47 x 0.47 x 1.55 mm , 256 x 256 x 60; sagittal
  • moving: subject 64 MRI: 0.55 x 0.55 x 1.5 mm , 256 x 256 x 60; sagittal

Download

Registration Results

rigidrigid
rigidrigid
affine alignmentaffine alignment
BSpline registration of full volumes. 9 x 9 x 5 gridBSpline registration of full volumes. 9 x 9 x 5 grid


Procedures

  • Phase 1: rigid alignment
  1. Go to the BRAINSfit module
    1. select Presets "Xf1_Rigid" or set the parameters as given below:
    2. fixed image: "58_MRI", moving image: "64_MRI"
    3. Initialize with previous transform: select "Off"
    4. Initialize Transform Mode: check box for use MomentsAlign
    5. check Include Rigid registration Phase box.
    6. Output: under Slicer Linear Transform, select new and rename to "Xf1_Rigid" or similar
    7. Registration Parameters: set "Number of Samples" to 150,000 at least
    8. leave rest at defaults
    9. Click Apply
  • Phase 2: affine alignment
  1. Go to the BRAINSfit module
    1. same as Phase 1 above, except:
    2. Initialize with previous transform: select "Xf1_Rigid" from phase 1 above
    3. Initialize Transform Mode: check box for Off
    4. check Include ScaleVersor" , "Include ScaleSkewVersor" and "Include Affine" registration Phase box.
    5. Registration Parameters: set "Number of Samples" to 200,000 at least
    6. Output: under Slicer Linear Transform, select new and rename to "Xf2_Affine" or similar
  • Phase 3: Masking
  1. Before allowing nonrigid deformations we must mask the aliasing present:
    1. Go to the Editor module.
    2. Select 58_MRI as Master Volume, choose the threshold tool and threshold at intensity ~68 to separate from the background.
    3. Select the Brush tool and cut connections between the aliasing portions and the rest of the image
    4. Select Island Tool and remove aliasing portions
    5. Select Dilate tool to fill holes
    6. Save. Repeat for the 64_MRI volume
  • Phase 4: BSpline alignment
  1. Go to the BRAINSfit module
    1. Initialize with previous transform: select "Xf2_Affine" from phase 2 above
    2. Output: under Slicer BSpline Transform, select new and rename to "Xf3_BSpline" or similar
    3. Output Image Volume: select new and rename to "64_MRI_Xf3" or similar
    4. Registration Parameters: set "Number of Samples" to 300,000 at least
    5. Number of Grid Subdivisions: 7,7,5
    6. Maximum B-Spline Displacement: set to 5 [mm]
    7. Masking: check ROI box, Input Fixed Mask: "58_MRI-label", Input Moving Mask: "64_MRI-label" generated in Phase 3 above.
    8. Click Apply
  2. Note to compare the different registrations: to see the Rigid or Affine registrations, drag the 64_MRI volume inside/onto the respective transform node. To see the BSpline registrations, select the "64_MRI_Xf3" volume as fore- or background.


Discussion: Registration Challenges

  • inter-subject registration: content differences are large, as are relative limb positions
  • aliasing: both MRIs have aliasing with part of the anatomy wrapping around to the other edge of the image. This needs to be masked for detailed registration

Discussion: Key Strategies

  • to calculate the transform, we use the images with the most accurate geometric representation and the smallest expected change, i.e. we align the follow-up CT to the baseline CT and then apply the transforms to the PET image.
  • because of the non-rigid differences due to posture and breathing we will need to apply a 2-step registration with an affine alignment followed by a BSpline.