Atlas file inputted into RABIES differs in orientation and fov to the output files of RABIES.

[KarisColyerPatel]

Hey All,

Using the output from the DR step as input for fsl randomise was effective! I am now trying to load the atlas (Fischer344) I inputted into RABIES, but there appears to be issues with the orientation and FOV (they differ between the files). Is there an output from the preprocessing step (of the atlas labels) that I could use to input into fsleyes. I checked the commonspace_labels but there are still issues with the FOV! Also the files contained here are per rat rather than averaged across my sample.

[Gab-D-G]

Hey Karis, the orientation is standardized to RAS during preprocessing, so this may differ from your inputs. Preprocessing will output a re-oriented/resampled version of the template and atlas files provided (there is commonspace_labels as you mentionned, but also commonspace_resampled_template for the resampled template). The resolution will be matching your functional images (or what was provided to --commonspace_resampling), as well as the FOV. Can you be more precise regarding the mismatch in FOV you observe with the commonspace_labels files? Is it relative to the commonspace preprocessed functional images?

[gdevenyi]

My guess here is that the input files are in non-standard orientation.

Can you open a representative set of input files (and the atlas) in itk-snap and share screenshots?

[KarisColyerPatel]

Of course! Yes, exactly the mismatch is relative to the commonspace preprocessed functional images. Please see below!
File outputted from DR (files are the commonspace preprocessed functional images)
filename sub-01_r1_a1_d1_b1_bold_combined_cleaned_DR_maps.nii.gz

sizeof_hdr 348
data_type FLOAT64
dim0 4
dim1 56
dim2 28
dim3 41
dim4 30
dim5 1
dim6 1
dim7 1
vox_units mm
time_units Unknown
datatype 64
nbyper 8
bitpix 64
pixdim0 1.000000
pixdim1 0.300000
pixdim2 1.000000
pixdim3 0.300000
pixdim4 0.801000
pixdim5 0.000000
pixdim6 0.000000
pixdim7 0.000000
vox_offset 352
cal_max 0.000000
cal_min 0.000000
scl_slope 1.000000
scl_inter 0.000000
phase_dim 0
freq_dim 0
slice_dim 0
slice_name Unknown
slice_code 0
slice_start 0
slice_end 0
slice_duration 0.000000
toffset 0.000000
intent Unknown
intent_code 0
intent_name
intent_p1 0.000000
intent_p2 0.000000
intent_p3 0.000000
qform_name Scanner Anat
qform_code 1
qto_xyz:1 0.300000 0.000000 0.000000 -9.140401
qto_xyz:2 0.000000 1.000000 0.000000 -16.190514
qto_xyz:3 0.000000 0.000000 0.300000 -0.845504
qto_xyz:4 0.000000 0.000000 0.000000 1.000000
qform_xorient Left-to-Right
qform_yorient Posterior-to-Anterior
qform_zorient Inferior-to-Superior
sform_name Scanner Anat
sform_code 0
sto_xyz:1 0.000000 0.000000 0.000000 0.000000
sto_xyz:2 0.000000 0.000000 0.000000 0.000000
sto_xyz:3 0.000000 0.000000 0.000000 0.000000
sto_xyz:4 0.000000 0.000000 0.000000 1.000000
sform_xorient Unknown
sform_yorient Unknown
sform_zorient Unknown
file_type NIFTI-1+
file_code 1
descrip

filename resampled_template.nii.gz

sizeof_hdr 348
data_type FLOAT32
dim0 3
dim1 68
dim2 114
dim3 50
dim4 1
dim5 0
dim6 0
dim7 0
vox_units mm
time_units Unknown
datatype 16
nbyper 4
bitpix 32
pixdim0 1.000000
pixdim1 0.250000
pixdim2 0.250000
pixdim3 0.250000
pixdim4 0.000000
pixdim5 0.000000
pixdim6 0.000000
pixdim7 0.000000
vox_offset 352
cal_max 0.000000
cal_min 0.000000
scl_slope 1.000000
scl_inter 0.000000
phase_dim 0
freq_dim 0
slice_dim 0
slice_name Unknown
slice_code 0
slice_start 0
slice_end 0
slice_duration 0.000000
toffset 0.000000
intent Unknown
intent_code 0
intent_name
intent_p1 0.000000
intent_p2 0.000000
intent_p3 0.000000
qform_name Scanner Anat
qform_code 1
qto_xyz:1 0.250000 0.000000 0.000000 -9.140400
qto_xyz:2 0.000000 0.250000 0.000000 -16.190500
qto_xyz:3 0.000000 0.000000 0.250000 -0.845504
qto_xyz:4 0.000000 0.000000 0.000000 1.000000
qform_xorient Left-to-Right
qform_yorient Posterior-to-Anterior
qform_zorient Inferior-to-Superior
sform_name Scanner Anat
sform_code 0
sto_xyz:1 0.000000 0.000000 0.000000 0.000000
sto_xyz:2 0.000000 0.000000 0.000000 0.000000
sto_xyz:3 0.000000 0.000000 0.000000 0.000000
sto_xyz:4 0.000000 0.000000 0.000000 1.000000
sform_xorient Unknown
sform_yorient Unknown
sform_zorient Unknown
file_type NIFTI-1+
file_code 1
descrip
aux_file

filename sub-01_r1_a1_d1_b1_bold_EPI_anat_labels.nii.gz

sizeof_hdr 348
data_type INT16
dim0 3
dim1 56
dim2 28
dim3 41
dim4 1
dim5 1
dim6 1
dim7 1
vox_units mm
time_units Unknown
datatype 4
nbyper 2
bitpix 16
pixdim0 1.000000
pixdim1 0.300000
pixdim2 1.000000
pixdim3 0.300000
pixdim4 0.000000
pixdim5 0.000000
pixdim6 0.000000
pixdim7 0.000000
vox_offset 352
cal_max 0.000000
cal_min 0.000000
scl_slope 1.000000
scl_inter 0.000000
phase_dim 0
freq_dim 0
slice_dim 0
slice_name Unknown
slice_code 0
slice_start 0
slice_end 0
slice_duration 0.000000
toffset 0.000000
intent Unknown
intent_code 0
intent_name
intent_p1 0.000000
intent_p2 0.000000
intent_p3 0.000000
qform_name Scanner Anat
qform_code 1
qto_xyz:1 0.300000 0.000000 0.000000 -9.140401
qto_xyz:2 0.000000 1.000000 0.000000 -16.190514
qto_xyz:3 0.000000 0.000000 0.300000 -0.845504
qto_xyz:4 0.000000 0.000000 0.000000 1.000000
qform_xorient Left-to-Right
qform_yorient Posterior-to-Anterior
qform_zorient Inferior-to-Superior
sform_name Scanner Anat
sform_code 1
sto_xyz:1 0.300000 0.000000 -0.000000 -9.140401
sto_xyz:2 0.000000 1.000000 -0.000000 -16.190514
sto_xyz:3 0.000000 0.000000 0.300000 -0.845504
sto_xyz:4 0.000000 0.000000 0.000000 1.000000
sform_xorient Left-to-Right
sform_yorient Posterior-to-Anterior
sform_zorient Inferior-to-Superior
file_type NIFTI-1+
file_code 1
descrip
aux_file

[gdevenyi]

The headers don't actually tell us what we need to know, which is the orientation of the data itself relative to the header coordinates interpreted by a tool we understand. Can you please produce some screenshots from ITK-snap. Thanks.

[KarisColyerPatel]

Ah yes I think you are right.

Although see below

The commonspace preprocessed functional data does appear to be registered to the resampled template and labels. But the warning "Displaying images with different orientations/fields of view!" is still shown in fsleyes. Would you guys suggest that I start from the top and sort out the differing orientations between the functional data and atlas?

[gdevenyi]

I'm sorry, I must ask again.

Please provide screenshots of representative input data, as well as the atlas you're using with ITK-snap.

[KarisColyerPatel]

As you can see when loading the two files, only one is visible at at time! This is the same if I try to overlay the files or have them side by side.

[KarisColyerPatel]

[Gab-D-G]

Hey, I'm not sure I fully understand. Did you check if the resampled_template file does overlap with the functional files? Can you show those side by side, and then show the resampled_template next to the Fischer template? The differences between the original Fischer template file and the resampled_template should be two-fold: a conversion to RAS orientation (this function

RABIES/rabies/preprocess_pkg/utils.py

Line 194 in a62094c

def convert_to_RAS(img_file, out_dir=None):

), and the resampling of the template resolution ot match that of you anatomical scans (the function here

RABIES/rabies/utils.py

Line 42 in a62094c

def resample_image_spacing(image, output_spacing):

). One of those two operations may cause the differences.

[gdevenyi]

The orientation of the input files is as expected. They are properly aligned with the anatomical directions.

@Gab-D-G perhaps you should be completely explicit on which output files are which, and what space they are expected to be in, after that @KarisColyerPatel can use ITK-snap to confirm said alignment.

[Gab-D-G]

Sorry if that wasn't clear. I would expect the resampled_template.nii.gz to be properly aligned with the preprocessed timeseries found in the commonspace_bold folder. We should confirm this is true. I don't see the overlap displayed in the previous comments. Then we can compare resampled_template.nii.gz to the original Fischer template to see where the mismatch originates from. Can you do this for us @KarisColyerPatel ?

I would suspect something to do with RAS orientation, but I may be wrong. In the resample_image_spacing function, this resampling operation is carried out https://github.com/CoBrALab/RABIES/blob/a62094cf0737b8241d70f73fb12aed215634891b/rabies/utils.py#L56C70-L56C70, which should in my understanding only change voxel size and change the matrix dimensions accordingly, but can you take a look @gdevenyi to confirm this is accurate?