Wrong nifti header when reading a nifti file using SimpleITK

[NirutaDhimal]

I am using Python 3.9, windows 10 OS and SimpleITK version 2.2.0. I used SimpleITK to load nifti image and view its header information. The pixdim[0] header is changed from -1 or 1 to 0. I used ITKsnap, nifti-tool and nibabel to see the header information and they all show -1 or 1.
I understand that this issue was recently fixed in ITK and this shouldn't be a problem in the version I am using.
https://github.com/InsightSoftwareConsortium/ITK/commit/9eb4c1209d215eed244e5e0a1d8f15942a61cb76#diff-fa33a46a6dc41cde5d18ee8f63cb941794512e03c6377340b8d110fe584cfdfa

I am willing to provide you the deidentified data. Please fix this or point me to a direction where I can fix this issue for my code.

This is the code that I used.
`brain_ncct_img = sitk.ReadImage(ncct_file)
brain_ncct_data = sitk.GetArrayFromImage(brain_ncct_img)

print("header using simpleitk")
for k in brain_ncct_img.GetMetaDataKeys():
v = brain_ncct_img.GetMetaData(k)
print("({0}) = = "{1}"".format(k,v))
`

This is the result I am getting.
header using simpleitk (ITK_FileNotes) = = "TE=3e+02;Time=172913.193" (ITK_original_direction) = = "[UNKNOWN_PRINT_CHARACTERISTICS] " (ITK_original_spacing) = = "[UNKNOWN_PRINT_CHARACTERISTICS] " (aux_file) = = "" (bitpix) = = "16" (cal_max) = = "0" (cal_min) = = "0" (datatype) = = "4" (descrip) = = "TE=3e+02;Time=172913.193" (dim[0]) = = "3" (dim[1]) = = "512" (dim[2]) = = "512" (dim[3]) = = "30" (dim[4]) = = "1" (dim[5]) = = "0" (dim[6]) = = "0" (dim[7]) = = "0" (dim_info) = = "0" (intent_code) = = "0" (intent_name) = = "" (intent_p1) = = "0" (intent_p2) = = "0" (intent_p3) = = "0" (nifti_type) = = "1" (pixdim[0]) = = "0" (pixdim[1]) = = "0.443359" (pixdim[2]) = = "0.443359" (pixdim[3]) = = "4.79859" (pixdim[4]) = = "0" (pixdim[5]) = = "0" (pixdim[6]) = = "0" (pixdim[7]) = = "0" (qform_code) = = "1" (qform_code_name) = = "NIFTI_XFORM_SCANNER_ANAT" (qoffset_x) = = "81.2783" (qoffset_y) = = "99.1478" (qoffset_z) = = "-85.4728" (quatern_b) = = "0" (quatern_c) = = "0.997859" (quatern_d) = = "0.0654031" (scl_inter) = = "-1024" (scl_slope) = = "1" (sform_code) = = "1" (sform_code_name) = = "NIFTI_XFORM_SCANNER_ANAT" (slice_code) = = "0" (slice_duration) = = "0" (slice_end) = = "0" (slice_start) = = "0" (srow_x) = = "-0.443359 0 -0 81.2783" (srow_y) = = "-0 0.439566 -0.626342 99.1478" (srow_z) = = "0 0.05787 4.75754 -85.4728" (toffset) = = "0" (vox_offset) = = "352" (xyzt_units) = = "10"

[dave3d]

I wrote a little test nifti generator:

import SimpleITK as sitk

img = sitk.GaussianSource(sitk.sitkUInt8, [100,100,100], [20,20,20], [50,50,50])
img.SetSpacing([-2, 3, 4])
print(img)
sitk.WriteImage(img, "test.nii.gz")

Then, when I load it and print out the metadata, I get this:

 pixdim[0] :  0
 pixdim[1] :  2
 pixdim[2] :  3
 pixdim[3] :  4

Looking at the Nifti spec (https://nifti.nimh.nih.gov/pub/dist/src/niftilib/nifti1.h), the pixel spacings start at pixdim[1], which is what I see and what your output shows.

[NirutaDhimal]

Can you try using this data and see the pixdim[0] value? The pixdim[1] value is correct and I am talking about pixdim[0]. why is it setting pixdim[0] as 0 when we clearly have its value as -1 in the nifti header? shouldn't it be -1?
https://drive.google.com/file/d/1Rmfkw7NmDh1Ljset9bp-xIFferZJcjNC/view?usp=share_link

[dave3d]

I've requested access to the file.

[dave3d]

So I wrote a little test program:

import nibabel as nib
import SimpleITK as sitk

# Load the initial file using NiBabel
filename = "1239-011_20220517172346_BL-NCCT.nii"
img = nib.load(filename)

# Print out pixdim
print("\nfile:", filename)
print("nibabel pixdim:", img.header['pixdim'])

# Load the file with SimpleITK and print pixdim
sitk_img = sitk.ReadImage(filename)
pixdim = []
for i in range(ord('0'), ord('8')):
    key = "pixdim[" + chr(i) + "]"
    pixdim.append( sitk_img.GetMetaData(key) )
print("\nSimpleITK pixdim:", pixdim)

# Write out a new Nifti file
sitk_filename = "1239-sitk.nii"
sitk.WriteImage(sitk_img, sitk_filename)

# Load the new file with NiBabel and print pixdim
img2 = nib.load(sitk_filename)
print("\nfile:", sitk_filename)
print("nibabel pixdim:", img2.header['pixdim'])

It does 3 things:

1. Load the file with NiBabel and print pixdim
2. Load the file with SimpleITK, print pixdim, write out a new Nifti file using SimpleITK
3. Load the new Nifti file with NiBabel and print pixdim

Here are the output results:

file: 1239-011_20220517172346_BL-NCCT.nii

nibabel pixdim: [-1.          0.44335938  0.44335938  4.7985935   0.          0.
  0.          0.        ]

SimpleITK pixdim: ['0', '0.443359', '0.443359', '4.79859', '0', '0', '0', '0']

file: 1239-sitk.nii
nibabel pixdim: [-1.          0.44335938  0.44335938  4.7985935   0.          0.
  0.          0.        ]

So it seems to me that, although SimpleITK doesn't keep pixdim[0] internally properly, when it writes out the Nifti, it is being set to the proper value. It seems that pixdim[0] has to do with how the quaternion transform is handled when loading the Nifti. But once SimpleITK has loaded the image, that doesn't really matter, since SimpleITK doesn't use that quaternion. Once SimpleITK writes out a new Nifti, the pixdim[0] value does seem to have the proper value.

[NirutaDhimal]

Thanks for helping me in this. I am creating a montage. I am reading the nifti file using simpleitk and using the pixel array to make the montage using matplotlib. So what's happening is if pixdim[0] is -1 the image is upside down so I will need to flip it upside down. I am currently using nibabel to get the pixdim[0] value. So my question is if the image has pixdim[0] value -1 or 1 and if simpleitk loads this image, shouldn't it have the same headers as that of the image? There may be case where we read a nifti file using simpleitk and perform some operations on the pixel data and save it in another format(not nifti). Can you fix this?

[dave3d]

Can you share with me your code? And any example dataset that is not being displayed in the proper orientation?

I can try and re-writing your montage code in SimpleITK and see if that fixes it.

Also, if you want to try another file format, I'd suggest NRRD. Nifti is an old, and rather odd format. NRRD is more straightforward and consistent.

[dave3d]

Another possibility is to try DICOMOrientImageFilter. Here's the documentation:

https://simpleitk.org/doxygen/latest/html/classitk_1_1simple_1_1DICOMOrientImageFilter.html