ANTS combine frames

Software Support
1

Hi there! I am new to ants and I wonder if anyone has encountered this little and probably very trivial issue.

Summary of what happened:

I am trying to combine motion correction and normalization for PET images as suggested here: Aligning PET data to MNI space using FMRIPREP transform. The idea is that this would reduce the number of times I need to reslice the data from 2 to 1. The code runs and I get a nice list of images, but I somehow can’t combine the images with ants.list_to_ndimage() .

The frames_in_MNI looks like this:

[ANTsImage (LPI)
 	 Pixel Type : float (float32)
 	 Components : 1
 	 Dimensions : (91, 109, 91)
 	 Spacing    : (2.0, 2.0, 2.0)
 	 Origin     : (90.0, 126.0, -72.0)
 	 Direction  : [-1.  0.  0.  0. -1.  0.  0.  0.  1.],
 ANTsImage (LPI)
 	 Pixel Type : float (float32)
 	 Components : 1
 	 Dimensions : (91, 109, 91)
 	 Spacing    : (2.0, 2.0, 2.0)
 	 Origin     : (90.0, 126.0, -72.0)
 	 Direction  : [-1.  0.  0.  0. -1.  0.  0.  0.  1.],
 ...

I have no idea what must give a origin value for each dimension (4) refers to. When doing motion correction as shown here: ANTsPy/tutorials/motionCorrectionExample.ipynb at master · ANTsX/ANTsPy · GitHub, the list motion_corrected also has only 3 values, eg. print(motion_corrected[0].origin)
(-358.8940124511719, -359.4070129394531, 128.38400268554688)

Thanks in advance!

Command used (and if a helper script was used, a link to the helper script or the command generated):


T1w2MNI = glob.glob(f'{bids_dir}/{subject}/{session}/anat/*MNI152NLin6Asym_mode-image_xfm.h5')[0]
PET2T1W =f'{regdir}/fwdtransform_PET2T1w_Affine.mat'
# motion_affines is a list of affines saved when motion correcting the PET 

images_unmerged = ants.ndimage_to_list( pet )
frames_in_MNI = list()

for i in range(len(images_unmerged)):
    comb_transforms = ants.apply_transforms(fixed=T1w_MNI,
                                        moving=images_unmerged[i],
                                        transformlist=[ T1w2MNI, PET2T1W, motion_affines[i]],
                                        compose=regdir)
    frame_MNI = ants.apply_transforms(fixed=T1w_MNI,moving=images_unmerged[i],transformlist=[comb_transforms])
    frames_in_MNI.append( frame_MNI )  
    

mc_MNI_PET = ants.list_to_ndimage( T1w_MNI, frames_in_MNI )

Version:

ants.version
‘0.5.3’

Relevant log outputs (up to 20 lines):

ValueError                                Traceback (most recent call last)
Cell In[77], line 12
      8     frame_MNI = ants.apply_transforms(fixed=T1w_MNI,moving=images_unmerged[i],transformlist=[comb_transforms])
      9     frames_in_MNI.append( frame_MNI )  
---> 12 mc_MNI_PET = ants.list_to_ndimage( T1w_MNI, frames_in_MNI )

File ~/.local/lib/python3.11/site-packages/ants/utils/ndimage_to_list.py:48, in list_to_ndimage(image, image_list)
     45         raise ValueError('all images must have the same pixeltype')
     47 dimensionout = ( *image_list[0].shape, len( image_list )  )
---> 48 newImage = ants.make_image(
     49   dimensionout,
     50   spacing = ants.get_spacing( image ),
     51   origin = ants.get_origin( image ),
     52   direction = ants.get_direction( image ),
     53   pixeltype = inpixeltype
     54   )
     55 # FIXME - should implement paste image filter from ITK
     56 for x in range( len( image_list ) ):

File ~/.local/lib/python3.11/site-packages/ants/core/ants_image_io.py:215, in make_image(imagesize, voxval, spacing, origin, direction, has_components, pixeltype)
    213 else:
    214     array = np.full(imagesize, voxval, dtype="float32")
--> 215 image = from_numpy(
    216     array,
...
    167 if len(new_origin) != self.dimension:
--> 168     raise ValueError('must give a origin value for each dimension (%i)' % self.dimension)
    170 get_lib_fn('setOrigin')(self.pointer, new_origin)

ValueError: must give a origin value for each dimension (4)

Screenshots / relevant information: