The voxel value is showing 'invalid'

Support
1

I have one image and one segmentation. When I load the image as volume and the segmentation as segmentation everything seems fine with both. There is no NAN voxel value in any of image or segmentation. I use the code below to mask the image with the segment. However, after that when i load the masked image I saw it as below in slicer in which the voxel value shows ‘invalid’ in the left down. why this happening and what does it mean?

inputpath='/home/ImgAve_RIPV_Aff.nii.gz'
newimg_path='/home/SegAve_RIPV_Aff_label_0.2.nii.gz'

img = sitk.ReadImage(inputpath)
img_arr=sitk.GetArrayFromImage(img)

seg = sitk.ReadImage(newimg_path)
seg_arr=sitk.GetArrayFromImage(seg)


img_arr[seg_arr==0]=0

newimg11 = sitk.GetImageFromArray(img_arr)

img_ave1 = copyImageParameters(newimg11,img, inputpath)
sitk.WriteImage(img_ave1,'/home/mmm.nii.gz')

unnamed
unnamed2100×1574 276 KB

2

I also checked that there is no NAN in none of the segment and image. The image is an average image of 50 cases.

3

There is nothing wrong. invalid color name just means that the label value 1332 has no color name specified in the chosen color table.

4

Thank you for your answer but the image which is called “ImgAve_RIPV_Aff_label_0.2” is not a label map to have a color map. It is average of 50 CT images with voxel values 0 to more than 1332. so, it is not like a label map with only one voxel value of 1332. It was a CT image which was masked by a segment with the python code and after being masked, ‘invalid’ appeared in the resultant ct masked image. I would be very grateful if you could help in clarifying.

5

The image appears in the label layer (indicated by the L in the Data Probe), therefore it is a labelmap volume.

I think there is a heuristics that by default an image that has label in the filename is loaded as a labelmap, so if you want to load this image as a scalar volume instead then you need to change the default in the “Add data” window (or if you load it using Python then specify the {"labelmap": 0} property input of slicer.util.loadVolume).

image