Hello!
I have two MRI images. One is 512x275x370 with 1x1x1 slice thickness and the second is 354x192x153 with a slice thickness of 1x1x2.6. My thinking is that I should resample the two images so that it has the same size (as the first MR image). Any idea to how that can be done?
thanks
As I see it you have two single files containing volumes...possibly a RAW file???? or DICOM???....my suggestion is to use either MATLAB to load the volume file modify it and then save it again in another file.....or even better use the software Voreen wich is a very powerful and free Volume Rendering Engine very easy to use and do the same thing load the volume file....divide it into single images, modify the image sizes in a loop and then save the new volume into another file
Firstly thanks for your reply
I will clarify the question :
from both MRI volumes:One is 512x275x370 with 1x1x1 slice thickness and the second is 354x192x153
I want transforer each in a new resolution 256*256*256 with a slice thickness of 1*1*1mm.
It seems that you have not only slightly different spatial resolutions, but also quite different fields of view (FOVs) of 51.2×27.5×37 cm3 and 35.4×19.2×39.78 cm3. Which means that you have to crop both volumes to the same final field of view of 25.6×25.6×25.6 cm3, i.e. you have to remove the outer parts of the volume (hopefully background or air, or at least irrelevant image data).
Before that, you probably should register one dataset on the other (depending on the kind of data, this may be done by pure translations, by rigid transformations including rotations, or by elastic registration - you could try e.g. a package like elastix (link below) or several others. If you are lucky, the registration tool also provides resampling to the same spatial resolution (voxel size). If not, you could use a simple one-dimensional linear interpolation of your second data set to interpolate the last dimension from 2.6 mm to 1.0 mm.
http://elastix.isi.uu.nl/
Just use Image J and change the canvas size of the second volume to match the matrix of the first volume in x and y. For the z direction, you will need to interpolate (reslice) the z direction from 2.6mm to 1mm resolution.
what we can use for python ? i have the same problem with CT scans any idea how to do it with python?
Many matlab or python functions can do image resize. Please be attention to the interpolation algorithm used in resizing. You should "guess" the intensity value of the inserted voxels. Commonly used interpolation methods include Nearest Interpolation, Bilinear Interpolation, Bicubic interpolation... It would be better to try them in several images and choose the best one. In many functions, you can simply change the function parameter to select different interpolation methods.
I know this is an old topic, however some readers might still benefit from this answer. SimpleITK handles this well. It has several wrappers, including Python.
Kamel Aloui You can use mrgrid command with regrid option from MRtrix3.
- Badges
- Science topic
- Similar topics
- Phytochemistry
- Extracts