In the dataset, the aneurysm location is provided as (x, y) coordinates. Since the imaging modalities are 3D (multiple slices), I’d like to clarify exactly which plane and which slice those coordinates refer to.
", "messages": [ { "id": 3290957, "postDate": "2025-09-18T18:06:34.353Z", "content": "In the dataset, the aneurysm location is provided as (x, y) coordinates. Since the imaging modalities are 3D (multiple slices), I’d like to clarify exactly which plane and which slice those coordinates refer to.
", "rawMarkdown": "In the dataset, the aneurysm location is provided as (x, y) coordinates. Since the imaging modalities are 3D (multiple slices), I’d like to clarify exactly which plane and which slice those coordinates refer to.", "votes": 1 }, { "id": 3291209, "postDate": "2025-09-19T07:29:28.623Z", "content": "The localizers table points to a specific SOPInstanceUID which is in most cases a .dcm file with a single slice. In the cases where there are multiple slices in a file, the coordinates in the localizers table have an additional field f, which is the slice index.
\nI made a simple notebook that visualizes the localizers. Have a look, it's relatively self-explanatory.
\nhttps://www.kaggle.com/code/thalro/check-localizers
Thank you for your comment. Is it correct that no one is responsible for processing the entire volume?
", "rawMarkdown": "Thank you for your comment. Is it correct that no one is responsible for processing the entire volume?", "replies": [ { "id": 3291507, "postDate": "2025-09-19T19:08:09.023Z", "content": "I‘m not sure what you mean. I certainly don‘t feel responsible for processing the entire volume :)
", "rawMarkdown": "I‘m not sure what you mean. I certainly don‘t feel responsible for processing the entire volume :)" }, { "id": 3291513, "postDate": "2025-09-19T19:54:16.477Z", "content": "In practice, we are dealing with a 3D volume rather than a single slice. However, in this dataset, only one slice per volume has been annotated. Since an aneurysm is inherently a volumetric structure, annotating just a single slice does not fully capture its spatial extent.
", "rawMarkdown": "In practice, we are dealing with a 3D volume rather than a single slice. However, in this dataset, only one slice per volume has been annotated. Since an aneurysm is inherently a volumetric structure, annotating just a single slice does not fully capture its spatial extent." }, { "id": 3291712, "postDate": "2025-09-20T07:00:04.583Z", "content": "I think that is part of the task of making his competition. You can look through some of the public notebooks and see how others do it.
", "rawMarkdown": "I think that is part of the task of making his competition. You can look through some of the public notebooks and see how others do it." }, { "id": 3293808, "postDate": "2025-09-24T17:19:33.827Z", "content": "The localizer is just a single point in roughly the center of the aneurysm. It does not represent the entirety of the aneurysm.
", "rawMarkdown": "The localizer is just a single point in roughly the center of the aneurysm. It does not represent the entirety of the aneurysm.", "votes": 1 } ] } ] } ], "comments": [ { "id": 3291209, "author_name": "thomas rost", "author_url": "", "post_date": "2025-09-19T07:29:28.623000", "content": "The localizers table points to a specific SOPInstanceUID which is in most cases a .dcm file with a single slice. In the cases where there are multiple slices in a file, the coordinates in the localizers table have an additional field f, which is the slice index.
\nI made a simple notebook that visualizes the localizers. Have a look, it's relatively self-explanatory.
\nhttps://www.kaggle.com/code/thalro/check-localizers
Thank you for your comment. Is it correct that no one is responsible for processing the entire volume?
", "votes": 0, "replies": [ { "id": 3291507, "author_name": "thomas rost", "author_url": "", "post_date": "2025-09-19T19:08:09.023000", "content": "I‘m not sure what you mean. I certainly don‘t feel responsible for processing the entire volume :)
", "votes": 0, "replies": [] }, { "id": 3291513, "author_name": "Taymaz R.Farshi", "author_url": "", "post_date": "2025-09-19T19:54:16.477000", "content": "In practice, we are dealing with a 3D volume rather than a single slice. However, in this dataset, only one slice per volume has been annotated. Since an aneurysm is inherently a volumetric structure, annotating just a single slice does not fully capture its spatial extent.
", "votes": 0, "replies": [] }, { "id": 3291712, "author_name": "thomas rost", "author_url": "", "post_date": "2025-09-20T07:00:04.583000", "content": "I think that is part of the task of making his competition. You can look through some of the public notebooks and see how others do it.
", "votes": 0, "replies": [] }, { "id": 3293808, "author_name": "Evan Calabrese", "author_url": "", "post_date": "2025-09-24T17:19:33.827000", "content": "The localizer is just a single point in roughly the center of the aneurysm. It does not represent the entirety of the aneurysm.
", "votes": 1, "replies": [] } ] } ] } ], "raw_markdown_by_id": { "3290957": "In the dataset, the aneurysm location is provided as (x, y) coordinates. Since the imaging modalities are 3D (multiple slices), I’d like to clarify exactly which plane and which slice those coordinates refer to.", "3291209": "The localizers table points to a specific SOPInstanceUID which is in most cases a .dcm file with a single slice. In the cases where there are multiple slices in a file, the coordinates in the localizers table have an additional field f, which is the slice index. \nI made a simple notebook that visualizes the localizers. Have a look, it's relatively self-explanatory.\nhttps://www.kaggle.com/code/thalro/check-localizers" } }