\n
![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F8734370%2F239ae26bca82c720d61a3e529e52680c%2FScreenshot%202025-08-25%20at%206.56.01AM.png?generation=1756119376761987&alt=media\")
Hi everyone, It looks like in the new version of the segmentation dataset, the simple .nii files and the _cowseg.nii files are no longer aligned. The shape, spacing and origin are now different per set, they were fine before update. Looking at this file 1.2.826.0.1.3680043.8.498.10035643165968342618460849823699311381_cowseg.nii I found Shape: (512, 512, 20), Voxels spacing: [0.4297 0.42970002 6.5 ]
\nImage origin: [ 99.58613586 -57.20310974 -51.74690247] and the corresponding file \"/kaggle/input/rsna-intracranial-aneurysm-detection/segmentations/1.2.826.0.1.3680043.8.498.10035643165968342618460849823699311381.nii\" has Shape: (512, 512, 228), Voxels dimension: [0.47460938 0.47460938 0.7000122 ]
\nImage origin: [128.26269531 111.23730469 864. ]. Are you experiencing the same problem?
I am not able to reproduce this issue. For 1.2.826.0.1.3680043.8.498.10035643165968342618460849823699311381, I get:
\nFor the image:
\nSpacing [0.474609, 0.474609, 0.700012]
\nOrigin [-128.263, -111.237, 864]
\nDirection
\n1 0 0
\n0 -1 0
\n0 0 1
\nImage Dimensions : [512, 512, 228]
\nBounding Box : {[-128.263 -111.237 864], [114.737 131.763 1023.6]}
And for the segmentation:
\nSpacing [0.474609, 0.474609, 0.700012]
\nOrigin [-128.263, -111.237, 864]
\nDirection
\n1 0 0
\n0 -1 0
\n0 0 1
\nImage Dimensions : [512, 512, 228]
\nBounding Box : {[-128.263 -111.237 864], [114.737 131.763 1023.6]}
Here is what I get when I load them in ITK-Snap I can see that they are correctly aligned:
\n
Can you let me know how you are loading your data and examining the header/shape? Perhaps that will reveal the issue.
", "rawMarkdown": "I am not able to reproduce this issue. For 1.2.826.0.1.3680043.8.498.10035643165968342618460849823699311381, I get:\n\nFor the image:\n\n```Spacing [0.474609, 0.474609, 0.700012]```\n```Origin [-128.263, -111.237, 864]```\n```Direction```\n```1 0 0```\n```0 -1 0```\n```0 0 1```\n```Image Dimensions : [512, 512, 228]```\n```Bounding Box : {[-128.263 -111.237 864], [114.737 131.763 1023.6]}```\n\nAnd for the segmentation:\n\n```Spacing [0.474609, 0.474609, 0.700012]```\n```Origin [-128.263, -111.237, 864]```\n```Direction```\n```1 0 0```\n```0 -1 0```\n```0 0 1```\n```Image Dimensions : [512, 512, 228]```\n```Bounding Box : {[-128.263 -111.237 864], [114.737 131.763 1023.6]}```\n\nHere is what I get when I load them in ITK-Snap I can see that they are correctly aligned:\n\n\nCan you let me know how you are loading your data and examining the header/shape? Perhaps that will reveal the issue.", "replies": [ { "id": 3274729, "postDate": "2025-08-25T11:14:11.467Z", "content": "Here is my code, beginner here :)
\nfile_path_1 = \"/kaggle/input/rsna-intracranial-aneurysm-detection/segmentations/1.2.826.0.1.3680043.8.498.10035643165968342618460849823699311381_cowseg.nii\"
cowseg_img = nib.load(file_path_1)
\ncowseg_data = cowseg_img.get_fdata()
\ncowseg_header = cowseg_img.header
\ncowseg_pixdim = cowseg_header['pixdim'][1:4]
cowseg_affine = cowseg_img.affine
\ncowseg_origin = cowseg_affine[:3, 3]
print(\"Shape:\", cowseg_data.shape)
\nprint(\"Data type:\", cowseg_data.dtype)
\nprint(\"Min value:\", np.min(cowseg_data))
\nprint(\"Max value:\", np.max(cowseg_data))
\nprint(\"Unique values:\", np.unique(cowseg_data))
print(\"Voxels spacing:\", cowseg_pixdim)
\nprint(\"Image origin: \", cowseg_origin)
simple_path = \"/kaggle/input/rsna-intracranial-aneurysm-detection/segmentations/1.2.826.0.1.3680043.8.498.10035643165968342618460849823699311381.nii\"
\nseg = nib.load(simple_path)
\nsimple_data = seg.get_fdata()
\nsimple_header = seg.header
\nsimple_pixdim = simple_header['pixdim'][1:4]
simple_affine = seg.affine
\nsimple_origin = simple_affine[:3, 3]
print(\"Shape:\", simple_data.shape)
\nprint(\"Data type:\", simple_data.dtype)
\nprint(\"Min value:\", np.min(simple_data))
\nprint(\"Max value:\", np.max(simple_data))
\nprint(\"Unique values:\", np.unique(simple_data))
print(\"Voxels dimension: \", simple_pixdim)
\nprint(\"Image origin:\", simple_origin)
It's working now. I don't know why but looks fine now. @evancalabrese, thank you for your response!
", "rawMarkdown": "It's working now. I don't know why but looks fine now. @evancalabrese, thank you for your response!" }, { "id": 3274810, "postDate": "2025-08-25T13:28:04.393Z", "content": "Thats great news! Thanks for the update.
", "rawMarkdown": "Thats great news! Thanks for the update." } ] } ] }, { "id": 3274607, "postDate": "2025-08-25T05:57:17.133Z", "content": "Hi everyone, It looks like in the new version of the segmentation dataset, the simple .nii files and the _cowseg.nii files are no longer aligned. The shape, spacing and origin are now different per set, they were fine before update. Looking at this file 1.2.826.0.1.3680043.8.498.10035643165968342618460849823699311381_cowseg.nii I found Shape: (512, 512, 20), Voxels spacing: [0.4297 0.42970002 6.5 ]
\nImage origin: [ 99.58613586 -57.20310974 -51.74690247] and the corresponding file \"/kaggle/input/rsna-intracranial-aneurysm-detection/segmentations/1.2.826.0.1.3680043.8.498.10035643165968342618460849823699311381.nii\" has Shape: (512, 512, 228), Voxels dimension: [0.47460938 0.47460938 0.7000122 ]
\nImage origin: [128.26269531 111.23730469 864. ]. Are you experiencing the same problem?
I am not able to reproduce this issue. For 1.2.826.0.1.3680043.8.498.10035643165968342618460849823699311381, I get:
\nFor the image:
\nSpacing [0.474609, 0.474609, 0.700012]
\nOrigin [-128.263, -111.237, 864]
\nDirection
\n1 0 0
\n0 -1 0
\n0 0 1
\nImage Dimensions : [512, 512, 228]
\nBounding Box : {[-128.263 -111.237 864], [114.737 131.763 1023.6]}
And for the segmentation:
\nSpacing [0.474609, 0.474609, 0.700012]
\nOrigin [-128.263, -111.237, 864]
\nDirection
\n1 0 0
\n0 -1 0
\n0 0 1
\nImage Dimensions : [512, 512, 228]
\nBounding Box : {[-128.263 -111.237 864], [114.737 131.763 1023.6]}
Here is what I get when I load them in ITK-Snap I can see that they are correctly aligned:
\n
Can you let me know how you are loading your data and examining the header/shape? Perhaps that will reveal the issue.
", "votes": 0, "replies": [ { "id": 3274729, "author_name": "CristinaDaniela Popa", "author_url": "", "post_date": "2025-08-25T11:14:11.467000", "content": "Here is my code, beginner here :)
\nfile_path_1 = \"/kaggle/input/rsna-intracranial-aneurysm-detection/segmentations/1.2.826.0.1.3680043.8.498.10035643165968342618460849823699311381_cowseg.nii\"
cowseg_img = nib.load(file_path_1)
\ncowseg_data = cowseg_img.get_fdata()
\ncowseg_header = cowseg_img.header
\ncowseg_pixdim = cowseg_header['pixdim'][1:4]
cowseg_affine = cowseg_img.affine
\ncowseg_origin = cowseg_affine[:3, 3]
print(\"Shape:\", cowseg_data.shape)
\nprint(\"Data type:\", cowseg_data.dtype)
\nprint(\"Min value:\", np.min(cowseg_data))
\nprint(\"Max value:\", np.max(cowseg_data))
\nprint(\"Unique values:\", np.unique(cowseg_data))
print(\"Voxels spacing:\", cowseg_pixdim)
\nprint(\"Image origin: \", cowseg_origin)
simple_path = \"/kaggle/input/rsna-intracranial-aneurysm-detection/segmentations/1.2.826.0.1.3680043.8.498.10035643165968342618460849823699311381.nii\"
\nseg = nib.load(simple_path)
\nsimple_data = seg.get_fdata()
\nsimple_header = seg.header
\nsimple_pixdim = simple_header['pixdim'][1:4]
simple_affine = seg.affine
\nsimple_origin = simple_affine[:3, 3]
print(\"Shape:\", simple_data.shape)
\nprint(\"Data type:\", simple_data.dtype)
\nprint(\"Min value:\", np.min(simple_data))
\nprint(\"Max value:\", np.max(simple_data))
\nprint(\"Unique values:\", np.unique(simple_data))
print(\"Voxels dimension: \", simple_pixdim)
\nprint(\"Image origin:\", simple_origin)
It's working now. I don't know why but looks fine now. @evancalabrese, thank you for your response!
", "votes": 0, "replies": [] }, { "id": 3274810, "author_name": "Evan Calabrese", "author_url": "", "post_date": "2025-08-25T13:28:04.393000", "content": "Thats great news! Thanks for the update.
", "votes": 0, "replies": [] } ] } ] } ], "raw_markdown_by_id": { "3274726": "I am not able to reproduce this issue. For 1.2.826.0.1.3680043.8.498.10035643165968342618460849823699311381, I get:\n\nFor the image:\n\n```Spacing [0.474609, 0.474609, 0.700012]```\n```Origin [-128.263, -111.237, 864]```\n```Direction```\n```1 0 0```\n```0 -1 0```\n```0 0 1```\n```Image Dimensions : [512, 512, 228]```\n```Bounding Box : {[-128.263 -111.237 864], [114.737 131.763 1023.6]}```\n\nAnd for the segmentation:\n\n```Spacing [0.474609, 0.474609, 0.700012]```\n```Origin [-128.263, -111.237, 864]```\n```Direction```\n```1 0 0```\n```0 -1 0```\n```0 0 1```\n```Image Dimensions : [512, 512, 228]```\n```Bounding Box : {[-128.263 -111.237 864], [114.737 131.763 1023.6]}```\n\nHere is what I get when I load them in ITK-Snap I can see that they are correctly aligned:\n\n\nCan you let me know how you are loading your data and examining the header/shape? Perhaps that will reveal the issue.", "3274607": "Hi everyone, It looks like in the new version of the segmentation dataset, the simple .nii files and the _cowseg.nii files are no longer aligned. The shape, spacing and origin are now different per set, they were fine before update. Looking at this file 1.2.826.0.1.3680043.8.498.10035643165968342618460849823699311381_cowseg.nii I found Shape: (512, 512, 20), Voxels spacing: [0.4297 0.42970002 6.5 ]\nImage origin: [ 99.58613586 -57.20310974 -51.74690247] and the corresponding file \"/kaggle/input/rsna-intracranial-aneurysm-detection/segmentations/1.2.826.0.1.3680043.8.498.10035643165968342618460849823699311381.nii\" has Shape: (512, 512, 228), Voxels dimension: [0.47460938 0.47460938 0.7000122 ]\nImage origin: [128.26269531 111.23730469 864. ]. Are you experiencing the same problem? " } }