![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F609324%2F7f7eeff905250e42a7417d95ed9baa9f%2Fimage.png?generation=1569095398908748&alt=media\")
I'm new to medical imaging, so sorry if this is a stupid question: can anyone say how to reconstruct the 3d structure of the CT scans? I.e. grouping all the dicoms and recombine them into a 514x514xS array, where S is the number of slices in the scan? Or is that impossible for this problem?
\n\nThanks!
", "messages": [ { "id": 631316, "postDate": "2019-09-21T19:48:59.203Z", "content": "You definitely can ! I reconstructed some ! \nFirst, you have to regroup together the slices with the same PatientID, then to sort them using ImagePositionPatient[2].\nSome examples using 3Dslicer (a free software) to visualize them:\n
.png?generation=1569095469630689&alt=media)
.png?generation=1569095485716864&alt=media)
I'm new to medical imaging, so sorry if this is a stupid question: can anyone say how to reconstruct the 3d structure of the CT scans? I.e. grouping all the dicoms and recombine them into a 514x514xS array, where S is the number of slices in the scan? Or is that impossible for this problem?
\n\nThanks!
", "rawMarkdown": "I'm new to medical imaging, so sorry if this is a stupid question: can anyone say how to reconstruct the 3d structure of the CT scans? I.e. grouping all the dicoms and recombine them into a 514x514xS array, where S is the number of slices in the scan? Or is that impossible for this problem?\n\nThanks!", "votes": 4 }, { "id": 630398, "postDate": "2019-09-20T07:12:24.617Z", "content": "this is possible but i'm not really sure if it's fine with this competition rule or not
", "rawMarkdown": "this is possible but i'm not really sure if it's fine with this competition rule or not", "votes": 2 }, { "id": 1749148, "postDate": "2022-04-08T08:56:30.167Z", "content": "If you ever read this, after regrouping slices with the same PatientID, better to sort them using SOPInstanceUID rather than position
", "rawMarkdown": "If you ever read this, after regrouping slices with the same PatientID, better to sort them using SOPInstanceUID rather than position" }, { "id": 630469, "postDate": "2019-09-20T08:50:01.383Z", "content": "You can probably use dicom metadata to group images by series ID and then reconstruct the scan. I'm trying it rn, but takes a long times to loop through all files.
", "rawMarkdown": "You can probably use dicom metadata to group images by series ID and then reconstruct the scan. I'm trying it rn, but takes a long times to loop through all files." } ], "comments": [ { "id": 631316, "author_name": "Al-Khwârizmî", "author_url": "", "post_date": "2019-09-21T19:48:59.203000", "content": "You definitely can ! I reconstructed some ! \nFirst, you have to regroup together the slices with the same PatientID, then to sort them using ImagePositionPatient[2].\nSome examples using 3Dslicer (a free software) to visualize them:\n
.png?generation=1569095469630689&alt=media)
.png?generation=1569095485716864&alt=media)
this is possible but i'm not really sure if it's fine with this competition rule or not
", "votes": 2, "replies": [] }, { "id": 1749148, "author_name": "Alexey Kostanov", "author_url": "", "post_date": "2022-04-08T08:56:30.167000", "content": "If you ever read this, after regrouping slices with the same PatientID, better to sort them using SOPInstanceUID rather than position
", "votes": 0, "replies": [] }, { "id": 630469, "author_name": "Robin Schwob", "author_url": "", "post_date": "2019-09-20T08:50:01.383000", "content": "You can probably use dicom metadata to group images by series ID and then reconstruct the scan. I'm trying it rn, but takes a long times to loop through all files.
", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "631316": "You definitely can ! I reconstructed some ! \nFirst, you have to regroup together the slices with the same PatientID, then to sort them using ImagePositionPatient[2].\nSome examples using 3Dslicer (a free software) to visualize them:\n\n\n.png?generation=1569095469630689&alt=media)\n\n.png?generation=1569095485716864&alt=media)\n", "630353": "I'm new to medical imaging, so sorry if this is a stupid question: can anyone say how to reconstruct the 3d structure of the CT scans? I.e. grouping all the dicoms and recombine them into a 514x514xS array, where S is the number of slices in the scan? Or is that impossible for this problem?\n\nThanks!", "630398": "this is possible but i'm not really sure if it's fine with this competition rule or not", "1749148": "If you ever read this, after regrouping slices with the same PatientID, better to sort them using SOPInstanceUID rather than position", "630469": "You can probably use dicom metadata to group images by series ID and then reconstruct the scan. I'm trying it rn, but takes a long times to loop through all files." } }