{ "id": 267714, "title": "Wrong DICOM metadata patient 108, T2w", "url": "/competitions/rsna-miccai-brain-tumor-radiogenomic-classification/discussion/267714", "author_name": "", "post_date": "2021-08-24T11:16:51.547689500Z", "votes": 6, "comment_count": 4, "views": 0, "content": "

@ujjwalbaid

\n

The T2w series of patient 108 seems to have incorrect metadata (duplication).
\nThe image position patient & slice position tags have been duplicated as I've highlighted in the image attached. Can you please check and confirm this ?

\n

\"https://ibb.co/MRrSvrw\"

", "messages": [ { "id": "1488550", "postDate": "08/24/2021 11:16:51", "content": "

@ujjwalbaid

\n

The T2w series of patient 108 seems to have incorrect metadata (duplication).
\nThe image position patient & slice position tags have been duplicated as I've highlighted in the image attached. Can you please check and confirm this ?

\n

\"https://ibb.co/MRrSvrw\"

", "rawMarkdown": "ujjwalbaid\n\nThe T2w series of patient 108 seems to have incorrect metadata (duplication). \nThe image position patient & slice position tags have been duplicated as I've highlighted in the image attached. Can you please check and confirm this ?\n\n![https://ibb.co/MRrSvrw] (https://ibb.co/MRrSvrw)", "votes": null }, { "id": "1488899", "postDate": "08/24/2021 15:17:18", "content": "

not exactly the same problem, but I ran into problems with this and a few other patients using image orientation based on ImageOrientationPatient, or using this code

", "rawMarkdown": "not exactly the same problem, but I ran into problems with this and a few other patients using image orientation based on ImageOrientationPatient, or using [this](https://www.kaggle.com/ren4yu/normalized-voxels-align-planes-and-crop) code", "votes": null }, { "id": "1489389", "postDate": "08/25/2021 00:30:26", "content": "

I use the same code with @yakovp. Here is a list of patients with image orientation and/or slice order problems I could find.

\n\n

I am so happy if anyone check this and share us the code to reconstruct correct 3D data.

", "rawMarkdown": "I use the same code with @yakovp. Here is a list of patients with image orientation and/or slice order problems I could find.\n- 108 (T2)\n- 137 (T1, T1wce, and FLAIR)\n- 148 (T1wce)\n\nI am so happy if anyone check this and share us the code to reconstruct correct 3D data.", "votes": null }, { "id": "1489943", "postDate": "08/25/2021 11:52:45", "content": "

you can find it that is in public notebooks.

", "rawMarkdown": "you can find it that is in public notebooks.", "votes": null }, { "id": "1489990", "postDate": "08/25/2021 12:32:58", "content": "

Would you tell me which public notebook?
\nI have tried these very popular notebooks. All of them did not worked for these patients.

\n", "rawMarkdown": "Would you tell me which public notebook?\nI have tried these very popular notebooks. All of them did not worked for these patients.\n- https://www.kaggle.com/ren4yu/normalized-voxels-align-planes-and-crop\n- https://www.kaggle.com/boojum/connecting-voxel-spaces\n- https://www.kaggle.com/rluethy/efficientnet3d-with-one-mri-type", "votes": null } ], "comments": [ { "id": 1488899, "author_name": "yakovp", "author_url": "", "post_date": "08/24/2021 15:17:18", "content": "

not exactly the same problem, but I ran into problems with this and a few other patients using image orientation based on ImageOrientationPatient, or using this code

", "votes": null, "replies": [ { "id": 1489389, "author_name": "tomooinubushi", "author_url": "", "post_date": "08/25/2021 00:30:26", "content": "

I use the same code with @yakovp. Here is a list of patients with image orientation and/or slice order problems I could find.

\n\n

I am so happy if anyone check this and share us the code to reconstruct correct 3D data.

", "votes": null, "replies": [] }, { "id": 1489943, "author_name": "dragonzhang", "author_url": "", "post_date": "08/25/2021 11:52:45", "content": "

you can find it that is in public notebooks.

", "votes": null, "replies": [] }, { "id": 1489990, "author_name": "tomooinubushi", "author_url": "", "post_date": "08/25/2021 12:32:58", "content": "

Would you tell me which public notebook?
\nI have tried these very popular notebooks. All of them did not worked for these patients.

\n", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "1488550": "ujjwalbaid\n\nThe T2w series of patient 108 seems to have incorrect metadata (duplication). \nThe image position patient & slice position tags have been duplicated as I've highlighted in the image attached. Can you please check and confirm this ?\n\n![https://ibb.co/MRrSvrw] (https://ibb.co/MRrSvrw)", "1488899": "not exactly the same problem, but I ran into problems with this and a few other patients using image orientation based on ImageOrientationPatient, or using [this](https://www.kaggle.com/ren4yu/normalized-voxels-align-planes-and-crop) code", "1489389": "I use the same code with @yakovp. Here is a list of patients with image orientation and/or slice order problems I could find.\n- 108 (T2)\n- 137 (T1, T1wce, and FLAIR)\n- 148 (T1wce)\n\nI am so happy if anyone check this and share us the code to reconstruct correct 3D data.", "1489943": "you can find it that is in public notebooks.", "1489990": "Would you tell me which public notebook?\nI have tried these very popular notebooks. All of them did not worked for these patients.\n- https://www.kaggle.com/ren4yu/normalized-voxels-align-planes-and-crop\n- https://www.kaggle.com/boojum/connecting-voxel-spaces\n- https://www.kaggle.com/rluethy/efficientnet3d-with-one-mri-type" }, "source": "meta" }