Documentation/Nightly/FAQ/Registration - Revision history

Grundlett: Text replacement - "https?:\/\/(?:www|wiki)\.slicer\.org\/slicerWiki\/index\.php\/([^ ]+) " to "https://www.slicer.org/wiki/$1 "

2019-11-27T13:20:24Z

Text replacement - "https?:\/\/(?:www|wiki)\.slicer\.org\/slicerWiki\/index\.php\/([^ ]+) " to "https://www.slicer.org/wiki/$1 "

Grundlett: Text replacement - "\[https:\/\/www\.slicer\.org\/slicerWiki\/index.php\/([^ ]+) ([^]]+)]" to "$2"

2019-11-27T02:40:12Z

Text replacement - "\[https:\/\/www\.slicer\.org\/slicerWiki\/index.php\/([^ ]+) ([^]]+)]" to "$2"

Grundlett: Text replacement - "\[http:\/\/www\.slicer\.org\/slicerWiki\/index\.php\/([^ ]+) ([^]]+)]" to "$2"

2019-11-27T02:36:54Z

Text replacement - "\[http:\/\/www\.slicer\.org\/slicerWiki\/index\.php\/([^ ]+) ([^]]+)]" to "$2"

Grundlett: Text replacement - "\[http:\/\/wiki\.slicer\.org\/slicerWiki\/index\.php\/([^ ]+) ([^]]+)]" to "$2"

2019-11-27T02:18:43Z

Text replacement - "\[http:\/\/wiki\.slicer\.org\/slicerWiki\/index\.php\/([^ ]+) ([^]]+)]" to "$2"

Inorton: /* The registration transform file saved by Slicer does not seem to match what is shown */

2018-09-25T19:28:55Z

The registration transform file saved by Slicer does not seem to match what is shown

Meier: /* Apply, Resample, Export */

2016-04-05T07:14:42Z

Apply, Resample, Export

Lasso: /* Is there a module for surface registration? */

2016-04-04T15:22:31Z

Is there a module for surface registration?

Meier: /* Is there a module for surface registration? */

2016-04-04T05:52:51Z

Is there a module for surface registration?

Meier: /* Parameters & Concept */

2016-04-02T20:21:52Z

Parameters & Concept

Meier: /* Where can I check the reference color map for the transform visualizer? i.e. how can I know the magnitude of transform vectors? */

2015-12-03T12:47:25Z

Where can I check the reference color map for the transform visualizer? i.e. how can I know the magnitude of transform vectors?

@@ Line 454: / Line 454: @@
 '''2.Automated registration:''' the basic concept here is to pad the image with duplicated slices in the 3rd dimension to provide the appearance of a 3D object (e.g. using the Resample module), and  then try to lock all degrees of freedom pertaining to the 3rd dimension (e.g. by masking), i.e. out-of-plane rotation or translation. Because the amount of data is very limited in 2D, automated registration is likely going to fail if the images are too far apart or to dissimilar in content.
 An example of 2D registration can be found in our case library:
-http://wiki.slicer.org/slicerWiki/index.php/Documentation:Nightly:Registration:RegistrationLibrary:RegLib_C46
+https://www.slicer.org/wiki/Documentation:Nightly:Registration:RegistrationLibrary:RegLib_C46 === What's the BSpline Grid Size? ===
 The BSpline grid is the basic structure of the deformation model applied when transforming one image to match the other. A uniform grid is placed over the entire image and the grid-points are allowed to move freely. These node displacements are then interpolated over the remainder of the grid. The image below shows an example of a 3x3x3 grid. An additional row of grid points is created at the boundary of the image (red). Those points are locked and do not move. They provide a boundary constraint to keep the grid from shifting. The interior points (green) are free to move in any direction (blue arrows), which gives 3 degrees of freedom (DOF) per grid-point. So a 3x3x3 grid translates into a transform with 27 DOF, a 5x5x5 grid will have 125 DOF. For comparison a rigid transform has 6 DOF, an affine transform has 12.
 When specifying a grid, be aware of your image resolution and anisotropy. For example if your image has substantially fewer voxels in one direction, it may be wise to also request fewer grid-points in that dimension, to avoid over fitting.<br>

@@ Line 575: / Line 575: @@
   referenceVolumeNode=slicer.util.getNode('MRHead')
   slicer.modules.transforms.logic().CreateDisplacementVolumeFromTransform(transformNode, referenceVolumeNode, False)
-The new volume can then be saved or used for ROI analysis with the [https://www.slicer.org/slicerWiki/index.php/Documentation/Nightly/Modules/LabelStatistics Label Statistics module].
+The new volume can then be saved or used for ROI analysis with the [[Documentation/Nightly/Modules/LabelStatistics|Label Statistics module]].
 === Physical Space vs. Image Space: how do I align two registered images to the same image grid? ===

@@ Line 559: / Line 559: @@
 Slice views: Slicer sets the overall field of view based on the image selected as "background". If the image in the foreground extends beyond that region it will be clipped. You can easily fix/test that by switching foreground and background volumes.
 If you already created a resampled version of the registered image, that may also have been cropped,because it sets the field of view based on the reference image (which is the fixed image usually). If you place your resampled image in the background and still see cropped edges, then that's what happened. In that case either use a different reference image for resampling or no reference at all and specify the result sizes manually, or use the  crop tool to expand the FOV as described below:
-#use the [http://www.slicer.org/slicerWiki/index.php/Documentation/Nightly/Modules/Crop_Volume CropVolume module] to over-crop the reference image (define ROI to cover the field of view you want to have in the resampled volume). You will need to use interpolated crop mode, so the resolution will not match the original reference volume.
+#use the [[Documentation/Nightly/Modules/Crop_Volume|CropVolume module]] to over-crop the reference image (define ROI to cover the field of view you want to have in the resampled volume). You will need to use interpolated crop mode, so the resolution will not match the original reference volume.
-#Use the result of the cropping operation as the reference for [http://www.slicer.org/slicerWiki/index.php/Documentation/Nightly/Modules/BRAINSResample BRAINSResample module] (you will need to specify the moving volume, output volume, and the transform produced by the BRAINS registration module).
+#Use the result of the cropping operation as the reference for [[Documentation/Nightly/Modules/BRAINSResample|BRAINSResample module]] (you will need to specify the moving volume, output volume, and the transform produced by the BRAINS registration module).

@@ Line 275: / Line 275: @@
 === How do I generate a mask and what should it look like ? ===
 The purpose of masking (see above) is to exclude any parts of the image that are present in only one of the images yet large enough to distract the registration algorithm as it is trying to find a match for something that doesn't exist. So your mask should focus on excluding those (irrelevant) aspects of the image. Also if you have deformations in the image and care about accuracy in one region more than another you would center the mask around that region. There is a set of screencasts dedicated to quick mask building here:
-[http://wiki.slicer.org/slicerWiki/index.php/Documentation/Nightly/RegistrationVideoTutorials Registration Video Tutorials]
+[[Documentation/Nightly/RegistrationVideoTutorials|Registration Video Tutorials]]
 <br>
 Basically you can use any segmentation tool in Slicer that produces a labelmap.
@@ Line 285: / Line 285: @@
 * you need a mask for both the fixed and the moving image.
-There are many example cases in the [http://wiki.slicer.org/slicerWiki/index.php/Documentation:Nightly:Registration:RegistrationLibrary '''Slicer Registration Library'''] that demonstrate the use of masks, e.g.:
+There are many example cases in the [[Documentation:Nightly:Registration:RegistrationLibrary|'''Slicer Registration Library''']] that demonstrate the use of masks, e.g.:
-*[http://wiki.slicer.org/slicerWiki/index.php/Documentation:Nightly:Registration:RegistrationLibrary:RegLib_C06 Case 06 (Breast MRI)]
+*[[Documentation:Nightly:Registration:RegistrationLibrary:RegLib_C06|Case 06 (Breast MRI)]]
-*[http://wiki.slicer.org/slicerWiki/index.php/Documentation:Nightly:Registration:RegistrationLibrary:RegLib_C07 Case 07 (prostate MRI)]
+*[[Documentation:Nightly:Registration:RegistrationLibrary:RegLib_C07|Case 07 (prostate MRI)]]
-*[http://wiki.slicer.org/slicerWiki/index.php/Documentation:Nightly:Registration:RegistrationLibrary:RegLib_C12 Case 12 (abdominal CT)]
+*[[Documentation:Nightly:Registration:RegistrationLibrary:RegLib_C12|Case 12 (abdominal CT)]]
-*[http://wiki.slicer.org/slicerWiki/index.php/Documentation:Nightly:Registration:RegistrationLibrary:RegLib_C17 Case 17 (interventional MRI-CT)]
+*[[Documentation:Nightly:Registration:RegistrationLibrary:RegLib_C17|Case 17 (interventional MRI-CT)]]
 === Which registration methods offer masking? ===
@@ Line 450: / Line 450: @@
 Slicer is designed to work with 3D images. Since it operates in physical space, all objects must have a physical dimension (e.g. thickness).
 It is possible to still run a 2D registration with the 3D tools, with limitations. The two options are: <br>
-'''1.manual registration:''' the manual registration can easily be restricted to 2D. Simply do not use the sliders that control out-of plane rotation or translation. For more info on manual registration see  [http://wiki.slicer.org/slicerWiki/index.php/Documentation/Nightly/FAQ#Can_I_manually_adjust_or_correct_a_registration.3F '''FAQ on manual registration'''] and also [http://wiki.slicer.org/slicerWiki/index.php/Documentation/Nightly/Registration/RegistrationLibrary the '''Registration Library''']
+'''1.manual registration:''' the manual registration can easily be restricted to 2D. Simply do not use the sliders that control out-of plane rotation or translation. For more info on manual registration see  [[Documentation/Nightly/FAQ#Can_I_manually_adjust_or_correct_a_registration.3F|'''FAQ on manual registration''']] and also [[Documentation/Nightly/Registration/RegistrationLibrary|the '''Registration Library''']]
 <br>
 '''2.Automated registration:''' the basic concept here is to pad the image with duplicated slices in the 3rd dimension to provide the appearance of a 3D object (e.g. using the Resample module), and  then try to lock all degrees of freedom pertaining to the 3rd dimension (e.g. by masking), i.e. out-of-plane rotation or translation. Because the amount of data is very limited in 2D, automated registration is likely going to fail if the images are too far apart or to dissimilar in content.
@@ Line 495: / Line 495: @@
 === How do I register images via landmarks/fiducials? ===
-There are two fiducial registration tools available within Slicer. The default is within the Registration menu, and an alternative version is found in the SlicerIGT extension as "Fiducial Registration Wizard". A detailed case example using both tools is available in the [http://wiki.slicer.org/slicerWiki/index.php/Documentation:Nightly:Registration:RegistrationLibrary:RegLib_C48 Case 48 in the Registration Case Library]. A brief screencast on selecting landmarks and running either tools are also here:
+There are two fiducial registration tools available within Slicer. The default is within the Registration menu, and an alternative version is found in the SlicerIGT extension as "Fiducial Registration Wizard". A detailed case example using both tools is available in the [[Documentation:Nightly:Registration:RegistrationLibrary:RegLib_C48|Case 48 in the Registration Case Library]]. A brief screencast on selecting landmarks and running either tools are also here:
 #[[Media:RegLib_C48_FiducialRegistration.mov|'''fiducial-based registration''']]
 #[[Media:RegLib_C48_FiducialRegistration_IGT.mov|'''alternative tool: using SlicerIGT's fiducial-based registration''']] (shows extension download and module execution only, for the landmark selection see movie above)

@@ Line 98: / Line 98: @@
 # Save the transform as a .tfm file.
 # Inspect the contents of the .tfm file in a text editor, and compare them to what is shown in the 4x4 matrix in the Transforms module.
-# re-load the .tfm back into slicer and confirm you have the same data as you saved from slicer.
+# re-load the .tfm back into slicer and confirm you have the same data as you saved from Slicer.
-you will notice that, even though the reloaded transform does match, the contents of the .tfm file and what is displayed in the Transforms module do not match. The explanation of this is kind of buried in the details of transforms. The issue relates to the difference between slicer which uses a "computer graphics" view of the world and itk which uses an "image processing" view of the world. By this we mean that in slicer you have a matrix hierarchy and you think in terms of moving an object from one spot to another - so a transform that has a positive "superior" value wrapped around a volume moves the volume up in patient space.<br>
-But ITK thinks of transformations in terms of mapping backwards from the display space back to the original image.  Imagine if you are stepping sequentially through the output pixels then ITK wants to know the transform that takes you back to the input pixels that it needs to use to calculate the output. This modeling vs. resampling issue is in addition to the LPS/RAS issue, which is the 2nd (invisible) difference between the two transforms.<br>
+You will notice that the original and re-loaded Transforms are identical, but do not match the content of the .tfm file. The issue relates to the difference between Slicer, which uses a "computer graphics" view of the world, and ITK, which uses an "image processing" view of the world. The Slicer transform hierarchy models movement of an object from one spot to another. For example, a transform that has a positive "superior" value wrapped around a volume moves the volume up in patient space.<br>
-In Summarry:
 #The transform represented in the widget is in RAS.
 #The transform represented in the tfm file is in LPS.
-#The transform represented in the file is the inverse of the transform in the widget (plus it has the LPS/RAS conversion applied).
+#The transform represented in the file is the inverse of the transform in the widget, and has the LPS/RAS conversion applied.
 #The order of the parameters in the tfm are the elements of the upper 3x3 of the transform displayed in the widget followed by the elements in the last column of the widget.
-As an example:
-# take the transform from the widget: --> c = [0.996918 -0.078459 -0.000000 6.899965; 0.068016 0.864225 -0.498488 -95.999726; 0.039111 0.496951 0.866896 266.299559; 0.0 0.000000 -0.000000 1.000000]
+Please see this discussion for more information, and code examples:
-#Take the inverse --> inv(c)  =
-.9969    0.0680    0.0391  -10.7644
+https://www.slicer.org/wiki/Documentation/Nightly/Modules/Transforms#Transform_files
 === I don't understand your coordinate system. What do the coordinate labels R,A,S and (negative numbers) mean? ===

@@ Line 571: / Line 571: @@
 Another alternative is to convert the transform into a 4-D deformation field directly and visualize it in slicer using RGB color. [[Documentation/{{documentation/version}}/FAQ#How_can_I_convert_a_BSpline_transform_into_a_deformation_field.3F|See FAQ below on how to convert.]]
 === Physical Space vs. Image Space: how do I align two registered images to the same image grid? ===
 Slicer displays all data in a physical coordinate system. Hence an image can only be displayed correctly if it contains sufficient header information to relate the image voxel grid with physical space. This includes voxel size, axis orientation and scan order.  It is therefore possible for two images to be aligned when viewed in Slicer, even though their underlying image grid is oriented very differently. To match the two images in image as well as physical space, the abovementioned axis direction, voxel size and image grid orientation must match. The procedure will depend on the image data, but the main tools at your disposal are the [[Documentation/{{documentation/version}}/Modules/ResampleScalarVectorDWIVolume|''ResampleScalarVectorDWIVolume'' module]] and  [[Documentation/{{documentation/version}}/Modules/OrientImages|''OrientImages'' module]].
 == Nonrigid / BSpline Registration ==

@@ Line 514: / Line 514: @@
 === Is there a module for surface registration? ===
-The current (4.5) version does not include a surface registration, but earlier Slicer versions (3.6 and before) include a surface registration module, you can get earlier versions here:
+The current (4.5) Slicer core does not include a surface registration, but available in extensions.
-https://www.slicer.org/pages/Special:SlicerDownloads
+* Rigid, similarity or affine registration of model to model: SlicerIGT extension, Model registration module (in IGT category)
-There also is a rigid ICP registration module for Slicer4 available as source. It requires a small simplification to register two models instead a model and markups:
+* Rigid, similarity or affine registration of markups fiducials to model: SlicerIGT extension, Fiducials-Model registration module (in IGT category)
 === Can I combine image and surface registration? ===

@@ Line 512: / Line 512: @@
 Yes, you can nest multiple (affine) registrations inside eachother. You can generate combined ones via the right-click context menu in the ''Data'' module and selecting ''Harden Transform''. Note that the original transform or volume is replaced when selecting ''Harden Transform'', so it is recommended to rename the node afterwards to document the fact. <br>
 Currently (v.4.1) BSpline transforms cannot be combined with other transforms, so if you have combinations of Affine and BSpline or multiple BSpline transforms you need to resample multiple times to apply them all. We recommend to supersample the volume beforehand to counteract interpolation blurring.
 === Can I combine image and surface registration? ===

@@ Line 593: / Line 593: @@
    /Applications/Slicer3.6.3/Slicer3 --launch /Applications/Slicer3.6.3/lib/Slicer3/Plugins/BSplineToDeformationField --help
-== Where can I check the reference color map for the transform visualizer? i.e. how can I know the magnitude of transform vectors? ==
+=== Where can I check the reference color map for the transform visualizer? i.e. how can I know the magnitude of transform vectors? ===
 The "Transform Visualizer" is a tool that in earlier Slicer versions (4.2 and before) was an extension module, but in newer versions has been integrated into the  [[Documentation/{{documentation/version}}/Modules/Transforms|'''Transforms''' module]]] (under the ''Display'' tab). It enables visualisation of transform vectors that are particularly useful to help understand complex non-rigid deformations such as B-Spline registrations.
 You can find the referenced color table in the [[Documentation/{{documentation/version}}/Modules/Colors|'''Colors''' module]] . There is a category called “Transform Display” and under that there is the “Displacement to color” color table. To display the scalar value represented by a color, you need to turn on the “Display scalar bar” option in Scalar Bar section under “Colors” module. Currently it only shows the scalar bar in the 3D window. <br>
 Alternatlively, you can see the deformation vector under current cursor in Transform module in the Information section, when you move your cursor, it will display the current vector.