|Please use this identifier to cite or link to this publication: http://hdl.handle.net/1926/563|
This filter offers a simple and intuitive way of extracting MPR's from volumetric datasets.Open Science:
All requirements of Open Science are met. Source code & tests are provided.Reproducibility:
I downloaded the code and the tests executed as expected.
Use of Open Source Software:
yes.Open source Contributions:
Yes. Code is clearly documented.Code Quality :
adheres to ITK standards
Quality of the data :
no data supplied.Interest:
This would be a quite useful class for any application that has to generate MPR's.Free comment :
I'd recommend adding a check for a pure rotation matrix.
Checking det(RotationMatrix) = 1 (pure rotation) or -1 (reflection) would be of use when the transform is used in registration algorithms.
The paper describes a method to specify a transform given the center point of a section, the section normal and the output size. Then it incorporates this transformation into a filter in order to obtain a 2D section image from a 3D image.
The author demonstrates the efficiency of the method by providing some examples and tests with synthetic and real data.
The work adheres to the concept of Open Science. Source code, test images and images shown in the paper are provided.
I managed to compile the source code and run the tests. Results were reproduced without problem as some tests with parameters were available.
Use of Open Source Software:
The implementation uses and extends ITK.
Open Source Contributions:
Source code is provided and its usage is straightforward.
In my opinion it is better to create a LookUpTransform that inherits from Rigid3DTransform instead of using initializer. I did this and incorporated some other changes (see attachment). These include:
- LookAtTransformInitializer changed to LookAtTransform that inherits Rigid3DTransform.
- m_Image is no longer a member as it is unnecessary just to specify output spacing. This spacing is incorporated as argument to the call SetPlane().
- The output spacing does not need to be the same as the input image spacing. Changed this (see previous).
- The template parameter for the transform is the 2D image, not the 3D image. This makes more sense in my opinion.
With respects to the tests, there is no usage information for command line as in ITK examples.
Applicability to other problems:
The method can be applied to any problem that requires the extraction of a section given its center and normal.
Suggestions for future work:
Requests for additional information from authors:
There is an error in the figure description. It says first and second row where it should say first and second column.
|Categories:||Filtering, Geometric transforms, Resampling|
|Keywords:||resample, ITK, oblique section, plane|
Linked Publications more
Anisotropic Fast Marching in ITK
by Mirebeau J.
ITK in Biomedical Research and Commercial Applications
by McCormick M., Aylward S., Johnson H., Lowekamp B.
Send a message to the author