{ "id": 460833, "title": "Important Information about the Hidden (Test) Datasets", "url": "/competitions/blood-vessel-segmentation/discussion/460833", "author_name": "", "post_date": "2023-12-11T11:14:53.422363100Z", "votes": 43, "comment_count": 11, "views": 0, "content": "

Hi All Kagglers,
\n It's been great watching the progress these last few weeks. Seeing a few things in the competition and also following the discussion carefully we decided that it makes sense to give some more information about the two hidden datasets i.e. the Public and Private test datasets:

\n

Public Test:
\nContinuous 3D part of a whole human kidney imaged with HiP-CT - Originally scanned at 25.14um/voxel and binned to 50.28um/voxel (bin x2) before segmentation.

\n

Private Test:
\nContinuous 3D part of a whole human kidney imaged with HiP-CT - Originally scanned at 15.77um/voxel binned to 63.08um/voxel (bin x4) before segmentation.

\n

Neither of these kidneys are different parts of ones already in the public dataset, i.e. they are physically different samples. The scanning and reconstruction procedures follow the method described in the original HiP-CT Methods paper (you should read this paper for background), and the manual annotation process is the same as in the benchmark review paper.

\n

We provided the higher resolution 5.2um/voxel VOI, to give some example data of different resolutions as we thought this might be useful.

", "messages": [ { "id": "2557282", "postDate": "12/11/2023 11:14:53", "content": "

Hi All Kagglers,
\n It's been great watching the progress these last few weeks. Seeing a few things in the competition and also following the discussion carefully we decided that it makes sense to give some more information about the two hidden datasets i.e. the Public and Private test datasets:

\n

Public Test:
\nContinuous 3D part of a whole human kidney imaged with HiP-CT - Originally scanned at 25.14um/voxel and binned to 50.28um/voxel (bin x2) before segmentation.

\n

Private Test:
\nContinuous 3D part of a whole human kidney imaged with HiP-CT - Originally scanned at 15.77um/voxel binned to 63.08um/voxel (bin x4) before segmentation.

\n

Neither of these kidneys are different parts of ones already in the public dataset, i.e. they are physically different samples. The scanning and reconstruction procedures follow the method described in the original HiP-CT Methods paper (you should read this paper for background), and the manual annotation process is the same as in the benchmark review paper.

\n

We provided the higher resolution 5.2um/voxel VOI, to give some example data of different resolutions as we thought this might be useful.

", "rawMarkdown": "Hi All Kagglers,\n It's been great watching the progress these last few weeks. Seeing a few things in the competition and also following the discussion carefully we decided that it makes sense to give some more information about the two hidden datasets i.e. the Public and Private test datasets: \n\nPublic Test: \nContinuous 3D part of a whole human kidney imaged with HiP-CT - Originally scanned at 25.14um/voxel and binned to 50.28um/voxel (bin x2) before segmentation. \n\nPrivate Test: \nContinuous 3D part of a whole human kidney imaged with HiP-CT - Originally scanned at 15.77um/voxel binned to 63.08um/voxel (bin x4) before segmentation. \n\n\nNeither of these kidneys are different parts of ones already in the public dataset, i.e. they are physically different samples. The scanning and reconstruction procedures follow the method described in the original HiP-CT [Methods paper](https://www.nature.com/articles/s41592-021-01317-x) (you should read this paper for background), and the manual annotation process is the same as in the benchmark [review paper](https://arxiv.org/pdf/2311.13319.pdf).\n\nWe provided the higher resolution 5.2um/voxel VOI, to give some example data of different resolutions as we thought this might be useful.", "votes": null }, { "id": "2557420", "postDate": "12/11/2023 13:30:39", "content": "

Thanks for the clarification!

", "rawMarkdown": "Thanks for the clarification!", "votes": null }, { "id": "2557491", "postDate": "12/11/2023 14:14:38", "content": "

i think you can also upate the data page. or provide a link to here:
\ne.g:

\n

\"\"

\n

sometimes, people don't read the forum

", "rawMarkdown": "i think you can also upate the data page. or provide a link to here:\ne.g:\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F113660%2F52bc463794ec6c35d065db3f30fc5850%2FSelection_999(4359).png?generation=1702304043847733&alt=media)\n\nsometimes, people don't read the forum", "votes": null }, { "id": "2557581", "postDate": "12/11/2023 15:10:33", "content": "

Thanks done!

", "rawMarkdown": "Thanks done!", "votes": null }, { "id": "2560195", "postDate": "12/13/2023 11:08:05", "content": "

Thank you for the helpful information!😀

\n

Regarding the private test set where it was mentioned that a whole human kidney was scanned at 15.77um/voxel, can I understand this as having the overall appearance of the entire kidney similar to the public set (like kidney_1_dense), but with just a difference in resolution? Or does it include only a part of the kidney, similar to kidney_1_voi, rather than encompassing the entire kidney?

", "rawMarkdown": "Thank you for the helpful information!😀\n\nRegarding the private test set where it was mentioned that a whole human kidney was scanned at 15.77um/voxel, can I understand this as having the overall appearance of the entire kidney similar to the public set (like kidney_1_dense), but with just a difference in resolution? Or does it include only a part of the kidney, similar to kidney_1_voi, rather than encompassing the entire kidney?", "votes": null }, { "id": "2567808", "postDate": "12/20/2023 01:42:34", "content": "

thank you for this important infomation.
\nIt's very helpfull.

", "rawMarkdown": "thank you for this important infomation.\nIt's very helpfull.", "votes": null }, { "id": "2599559", "postDate": "01/13/2024 00:05:36", "content": "

thank you

", "rawMarkdown": "thank you", "votes": null }, { "id": "2602153", "postDate": "01/14/2024 23:39:54", "content": "

Thanks.I find the reason of the lb score is zero.

", "rawMarkdown": "Thanks.I find the reason of the lb score is zero.", "votes": null }, { "id": "2605196", "postDate": "01/16/2024 23:16:56", "content": "

Thanks for the update!

", "rawMarkdown": "Thanks for the update!", "votes": null }, { "id": "2608636", "postDate": "01/19/2024 01:22:22", "content": "
\n

Hi All Kagglers,
\n It's been great watching the progress these last few weeks. Seeing a few things in the competition and also following the discussion carefully we decided that it makes sense to give some more information about the two hidden datasets i.e. the Public and Private test datasets:

\n

Public Test:
\n Continuous 3D part of a whole human kidney imaged with HiP-CT - Originally scanned at 25.14um/voxel and binned to 50.28um/voxel (bin x2) before segmentation.

\n

Private Test:
\n Continuous 3D part of a whole human kidney imaged with HiP-CT - Originally scanned at 15.77um/voxel binned to 63.08um/voxel (bin x4) before segmentation.

\n

Neither of these kidneys are different parts of ones already in the public dataset, i.e. they are physically different samples. The scanning and reconstruction procedures follow the method described in the original HiP-CT Methods paper (you should read this paper for background), and the manual annotation process is the same as in the benchmark review paper.

\n

We provided the higher resolution 5.2um/voxel VOI, to give some example data of different resolutions as we thought this might be useful.

\n

Hi All Kagglers,
\n It's been great watching the progress these last few weeks. Seeing a few things in the competition and also following the discussion carefully we decided that it makes sense to give some more information about the two hidden datasets i.e. the Public and Private test datasets:

\n

Public Test:
\n Continuous 3D part of a whole human kidney imaged with HiP-CT - Originally scanned at 25.14um/voxel and binned to 50.28um/voxel (bin x2) before segmentation.

\n

Private Test:
\n Continuous 3D part of a whole human kidney imaged with HiP-CT - Originally scanned at 15.77um/voxel binned to 63.08um/voxel (bin x4) before segmentation.

\n

Neither of these kidneys are different parts of ones already in the public dataset, i.e. they are physically different samples. The scanning and reconstruction procedures follow the method described in the original HiP-CT Methods paper (you should read this paper for background), and the manual annotation process is the same as in the benchmark review paper.

\n

We provided the higher resolution 5.2um/voxel VOI, to give some example data of different resolutions as we thought this might be useful.

\n
", "rawMarkdown": "> Hi All Kagglers,\n> It's been great watching the progress these last few weeks. Seeing a few things in the competition and also following the discussion carefully we decided that it makes sense to give some more information about the two hidden datasets i.e. the Public and Private test datasets: \n> \n> Public Test: \n> Continuous 3D part of a whole human kidney imaged with HiP-CT - Originally scanned at 25.14um/voxel and binned to 50.28um/voxel (bin x2) before segmentation. \n> \n> Private Test: \n> Continuous 3D part of a whole human kidney imaged with HiP-CT - Originally scanned at 15.77um/voxel binned to 63.08um/voxel (bin x4) before segmentation. \n> \n> \n> Neither of these kidneys are different parts of ones already in the public dataset, i.e. they are physically different samples. The scanning and reconstruction procedures follow the method described in the original HiP-CT [Methods paper](https://www.nature.com/articles/s41592-021-01317-x) (you should read this paper for background), and the manual annotation process is the same as in the benchmark [review paper](https://arxiv.org/pdf/2311.13319.pdf).\n> \n> We provided the higher resolution 5.2um/voxel VOI, to give some example data of different resolutions as we thought this might be useful.\n\n> Hi All Kagglers,\n> It's been great watching the progress these last few weeks. Seeing a few things in the competition and also following the discussion carefully we decided that it makes sense to give some more information about the two hidden datasets i.e. the Public and Private test datasets: \n> \n> Public Test: \n> Continuous 3D part of a whole human kidney imaged with HiP-CT - Originally scanned at 25.14um/voxel and binned to 50.28um/voxel (bin x2) before segmentation. \n> \n> Private Test: \n> Continuous 3D part of a whole human kidney imaged with HiP-CT - Originally scanned at 15.77um/voxel binned to 63.08um/voxel (bin x4) before segmentation. \n> \n> \n> Neither of these kidneys are different parts of ones already in the public dataset, i.e. they are physically different samples. The scanning and reconstruction procedures follow the method described in the original HiP-CT [Methods paper](https://www.nature.com/articles/s41592-021-01317-x) (you should read this paper for background), and the manual annotation process is the same as in the benchmark [review paper](https://arxiv.org/pdf/2311.13319.pdf).\n> \n> We provided the higher resolution 5.2um/voxel VOI, to give some example data of different resolutions as we thought this might be useful.", "votes": null }, { "id": "2608637", "postDate": "01/19/2024 01:22:39", "content": "

thank you

", "rawMarkdown": "thank you", "votes": null }, { "id": "2620557", "postDate": "01/26/2024 07:57:23", "content": "

Thanks for the info!

", "rawMarkdown": "Thanks for the info!", "votes": null } ], "comments": [ { "id": 2557420, "author_name": "hengck23", "author_url": "", "post_date": "12/11/2023 13:30:39", "content": "

Thanks for the clarification!

", "votes": null, "replies": [] }, { "id": 2557491, "author_name": "hengck23", "author_url": "", "post_date": "12/11/2023 14:14:38", "content": "

i think you can also upate the data page. or provide a link to here:
\ne.g:

\n

\"\"

\n

sometimes, people don't read the forum

", "votes": null, "replies": [ { "id": 2557581, "author_name": "clairewalsh", "author_url": "", "post_date": "12/11/2023 15:10:33", "content": "

Thanks done!

", "votes": null, "replies": [] } ] }, { "id": 2560195, "author_name": "siwooyong", "author_url": "", "post_date": "12/13/2023 11:08:05", "content": "

Thank you for the helpful information!😀

\n

Regarding the private test set where it was mentioned that a whole human kidney was scanned at 15.77um/voxel, can I understand this as having the overall appearance of the entire kidney similar to the public set (like kidney_1_dense), but with just a difference in resolution? Or does it include only a part of the kidney, similar to kidney_1_voi, rather than encompassing the entire kidney?

", "votes": null, "replies": [] }, { "id": 2567808, "author_name": "udaken10", "author_url": "", "post_date": "12/20/2023 01:42:34", "content": "

thank you for this important infomation.
\nIt's very helpfull.

", "votes": null, "replies": [] }, { "id": 2599559, "author_name": "kazukimorikawa", "author_url": "", "post_date": "01/13/2024 00:05:36", "content": "

thank you

", "votes": null, "replies": [] }, { "id": 2602153, "author_name": "mornicen", "author_url": "", "post_date": "01/14/2024 23:39:54", "content": "

Thanks.I find the reason of the lb score is zero.

", "votes": null, "replies": [] }, { "id": 2605196, "author_name": "syedabdul", "author_url": "", "post_date": "01/16/2024 23:16:56", "content": "

Thanks for the update!

", "votes": null, "replies": [] }, { "id": 2608636, "author_name": "fanzengming", "author_url": "", "post_date": "01/19/2024 01:22:22", "content": "
\n

Hi All Kagglers,
\n It's been great watching the progress these last few weeks. Seeing a few things in the competition and also following the discussion carefully we decided that it makes sense to give some more information about the two hidden datasets i.e. the Public and Private test datasets:

\n

Public Test:
\n Continuous 3D part of a whole human kidney imaged with HiP-CT - Originally scanned at 25.14um/voxel and binned to 50.28um/voxel (bin x2) before segmentation.

\n

Private Test:
\n Continuous 3D part of a whole human kidney imaged with HiP-CT - Originally scanned at 15.77um/voxel binned to 63.08um/voxel (bin x4) before segmentation.

\n

Neither of these kidneys are different parts of ones already in the public dataset, i.e. they are physically different samples. The scanning and reconstruction procedures follow the method described in the original HiP-CT Methods paper (you should read this paper for background), and the manual annotation process is the same as in the benchmark review paper.

\n

We provided the higher resolution 5.2um/voxel VOI, to give some example data of different resolutions as we thought this might be useful.

\n

Hi All Kagglers,
\n It's been great watching the progress these last few weeks. Seeing a few things in the competition and also following the discussion carefully we decided that it makes sense to give some more information about the two hidden datasets i.e. the Public and Private test datasets:

\n

Public Test:
\n Continuous 3D part of a whole human kidney imaged with HiP-CT - Originally scanned at 25.14um/voxel and binned to 50.28um/voxel (bin x2) before segmentation.

\n

Private Test:
\n Continuous 3D part of a whole human kidney imaged with HiP-CT - Originally scanned at 15.77um/voxel binned to 63.08um/voxel (bin x4) before segmentation.

\n

Neither of these kidneys are different parts of ones already in the public dataset, i.e. they are physically different samples. The scanning and reconstruction procedures follow the method described in the original HiP-CT Methods paper (you should read this paper for background), and the manual annotation process is the same as in the benchmark review paper.

\n

We provided the higher resolution 5.2um/voxel VOI, to give some example data of different resolutions as we thought this might be useful.

\n
", "votes": null, "replies": [ { "id": 2608637, "author_name": "fanzengming", "author_url": "", "post_date": "01/19/2024 01:22:39", "content": "

thank you

", "votes": null, "replies": [] } ] }, { "id": 2620557, "author_name": "arthurqzhao", "author_url": "", "post_date": "01/26/2024 07:57:23", "content": "

Thanks for the info!

", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "2557282": "Hi All Kagglers,\n It's been great watching the progress these last few weeks. Seeing a few things in the competition and also following the discussion carefully we decided that it makes sense to give some more information about the two hidden datasets i.e. the Public and Private test datasets: \n\nPublic Test: \nContinuous 3D part of a whole human kidney imaged with HiP-CT - Originally scanned at 25.14um/voxel and binned to 50.28um/voxel (bin x2) before segmentation. \n\nPrivate Test: \nContinuous 3D part of a whole human kidney imaged with HiP-CT - Originally scanned at 15.77um/voxel binned to 63.08um/voxel (bin x4) before segmentation. \n\n\nNeither of these kidneys are different parts of ones already in the public dataset, i.e. they are physically different samples. The scanning and reconstruction procedures follow the method described in the original HiP-CT [Methods paper](https://www.nature.com/articles/s41592-021-01317-x) (you should read this paper for background), and the manual annotation process is the same as in the benchmark [review paper](https://arxiv.org/pdf/2311.13319.pdf).\n\nWe provided the higher resolution 5.2um/voxel VOI, to give some example data of different resolutions as we thought this might be useful.", "2557420": "Thanks for the clarification!", "2557491": "i think you can also upate the data page. or provide a link to here:\ne.g:\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F113660%2F52bc463794ec6c35d065db3f30fc5850%2FSelection_999(4359).png?generation=1702304043847733&alt=media)\n\nsometimes, people don't read the forum", "2557581": "Thanks done!", "2560195": "Thank you for the helpful information!😀\n\nRegarding the private test set where it was mentioned that a whole human kidney was scanned at 15.77um/voxel, can I understand this as having the overall appearance of the entire kidney similar to the public set (like kidney_1_dense), but with just a difference in resolution? Or does it include only a part of the kidney, similar to kidney_1_voi, rather than encompassing the entire kidney?", "2567808": "thank you for this important infomation.\nIt's very helpfull.", "2599559": "thank you", "2602153": "Thanks.I find the reason of the lb score is zero.", "2605196": "Thanks for the update!", "2608636": "> Hi All Kagglers,\n> It's been great watching the progress these last few weeks. Seeing a few things in the competition and also following the discussion carefully we decided that it makes sense to give some more information about the two hidden datasets i.e. the Public and Private test datasets: \n> \n> Public Test: \n> Continuous 3D part of a whole human kidney imaged with HiP-CT - Originally scanned at 25.14um/voxel and binned to 50.28um/voxel (bin x2) before segmentation. \n> \n> Private Test: \n> Continuous 3D part of a whole human kidney imaged with HiP-CT - Originally scanned at 15.77um/voxel binned to 63.08um/voxel (bin x4) before segmentation. \n> \n> \n> Neither of these kidneys are different parts of ones already in the public dataset, i.e. they are physically different samples. The scanning and reconstruction procedures follow the method described in the original HiP-CT [Methods paper](https://www.nature.com/articles/s41592-021-01317-x) (you should read this paper for background), and the manual annotation process is the same as in the benchmark [review paper](https://arxiv.org/pdf/2311.13319.pdf).\n> \n> We provided the higher resolution 5.2um/voxel VOI, to give some example data of different resolutions as we thought this might be useful.\n\n> Hi All Kagglers,\n> It's been great watching the progress these last few weeks. Seeing a few things in the competition and also following the discussion carefully we decided that it makes sense to give some more information about the two hidden datasets i.e. the Public and Private test datasets: \n> \n> Public Test: \n> Continuous 3D part of a whole human kidney imaged with HiP-CT - Originally scanned at 25.14um/voxel and binned to 50.28um/voxel (bin x2) before segmentation. \n> \n> Private Test: \n> Continuous 3D part of a whole human kidney imaged with HiP-CT - Originally scanned at 15.77um/voxel binned to 63.08um/voxel (bin x4) before segmentation. \n> \n> \n> Neither of these kidneys are different parts of ones already in the public dataset, i.e. they are physically different samples. The scanning and reconstruction procedures follow the method described in the original HiP-CT [Methods paper](https://www.nature.com/articles/s41592-021-01317-x) (you should read this paper for background), and the manual annotation process is the same as in the benchmark [review paper](https://arxiv.org/pdf/2311.13319.pdf).\n> \n> We provided the higher resolution 5.2um/voxel VOI, to give some example data of different resolutions as we thought this might be useful.", "2608637": "thank you", "2620557": "Thanks for the info!" }, "source": "meta" }