{ "id": 475337, "title": "2024 paper automate the annotation task of 3d vasculature", "url": "/competitions/blood-vessel-segmentation/discussion/475337", "author_name": "", "post_date": "2024-02-08T04:08:50.528993400Z", "votes": 4, "comment_count": 1, "views": 0, "content": "

@clairewalsh
\nhttps://arxiv.org/pdf/2401.13961.pdf
\nTRISAM: TRI-PLANE SAM FOR ZERO-SHOT CORTICAL BLOOD VESSEL SEGMENTATION IN VEM IMAGES
\ni read the paper … basically the same process: seed in one of the 3 x,y,z plane, then floodfill using SAM
\n
\n
\n
\n
\nwonder if the oragnizer is keen to ask kaggler to devlope this?
\n\"We standardized the resolution, addressed imaging variations, and meticulously annotated blood vessels through semi-automatic,
\nmanual, and quality control processes, ensuring high-quality 3D segmentation. Furthermore, we developed a zero-shot cortical blood vessel segmentation method named TriSAM, which leverages the powerful segmentation model SAM for 3D
\nsegmentation. To lift SAM from 2D segmentation to 3D volume segmentation, TriSAM employs a multi-seed tracking framework, leveraging the reliability of certain image planes for tracking while using others to identify potential turning points. This approach, consisting of Tri-Plane selection, SAM-based tracking, and recursive redirection, effectively achieves long-term 3D blood vessel segmentation
\nwithout model training or fine-tuning. \"

", "messages": [ { "id": "2642264", "postDate": "02/08/2024 04:08:50", "content": "

@clairewalsh
\nhttps://arxiv.org/pdf/2401.13961.pdf
\nTRISAM: TRI-PLANE SAM FOR ZERO-SHOT CORTICAL BLOOD VESSEL SEGMENTATION IN VEM IMAGES
\ni read the paper … basically the same process: seed in one of the 3 x,y,z plane, then floodfill using SAM
\n
\n
\n
\n
\nwonder if the oragnizer is keen to ask kaggler to devlope this?
\n\"We standardized the resolution, addressed imaging variations, and meticulously annotated blood vessels through semi-automatic,
\nmanual, and quality control processes, ensuring high-quality 3D segmentation. Furthermore, we developed a zero-shot cortical blood vessel segmentation method named TriSAM, which leverages the powerful segmentation model SAM for 3D
\nsegmentation. To lift SAM from 2D segmentation to 3D volume segmentation, TriSAM employs a multi-seed tracking framework, leveraging the reliability of certain image planes for tracking while using others to identify potential turning points. This approach, consisting of Tri-Plane selection, SAM-based tracking, and recursive redirection, effectively achieves long-term 3D blood vessel segmentation
\nwithout model training or fine-tuning. \"

", "rawMarkdown": "@clairewalsh\nhttps://arxiv.org/pdf/2401.13961.pdf\nTRISAM: TRI-PLANE SAM FOR ZERO-SHOT CORTICAL BLOOD VESSEL SEGMENTATION IN VEM IMAGES\n\ni read the paper ... basically the same process: seed in one of the 3 x,y,z plane, then floodfill using SAM\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F113660%2F36ad01490d420c1876c21462341cef46%2FSelection_999(4958).png?generation=1707365281794226&alt=media)\n\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F113660%2F043438468c2d97bf3c06c13529f5d9db%2FSelection_999(4960).png?generation=1707365269017070&alt=media)\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F113660%2F33d6292ad97bebde2d406e169dee0d88%2FSelection_999(4962).png?generation=1707365593078826&alt=media)\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F113660%2Fef75d3d8291928dc770b3bd45791b842%2FSelection_999(4961).png?generation=1707365306046858&alt=media)\n\nwonder if the oragnizer is keen to ask kaggler to devlope this?\n\n\"We standardized the resolution, addressed imaging variations, and meticulously annotated blood vessels through semi-automatic,\nmanual, and quality control processes, ensuring high-quality 3D segmentation. Furthermore, we developed a zero-shot cortical blood vessel segmentation method named TriSAM, which leverages the powerful segmentation model SAM for 3D\nsegmentation. To lift SAM from 2D segmentation to 3D volume segmentation, TriSAM employs a multi-seed tracking framework, leveraging the reliability of certain image planes for tracking while using others to identify potential turning points. This approach, consisting of Tri-Plane selection, SAM-based tracking, and recursive redirection, effectively achieves long-term 3D blood vessel segmentation\nwithout model training or fine-tuning. \"", "votes": null }, { "id": "2642268", "postDate": "02/08/2024 04:13:14", "content": "

", "rawMarkdown": "![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F113660%2Fa344bc7d99b126bcdfb543ebb0e7c4f3%2FSelection_999(4963).png?generation=1707365581244185&alt=media)", "votes": null } ], "comments": [ { "id": 2642268, "author_name": "hengck23", "author_url": "", "post_date": "02/08/2024 04:13:14", "content": "

", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "2642264": "@clairewalsh\nhttps://arxiv.org/pdf/2401.13961.pdf\nTRISAM: TRI-PLANE SAM FOR ZERO-SHOT CORTICAL BLOOD VESSEL SEGMENTATION IN VEM IMAGES\n\ni read the paper ... basically the same process: seed in one of the 3 x,y,z plane, then floodfill using SAM\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F113660%2F36ad01490d420c1876c21462341cef46%2FSelection_999(4958).png?generation=1707365281794226&alt=media)\n\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F113660%2F043438468c2d97bf3c06c13529f5d9db%2FSelection_999(4960).png?generation=1707365269017070&alt=media)\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F113660%2F33d6292ad97bebde2d406e169dee0d88%2FSelection_999(4962).png?generation=1707365593078826&alt=media)\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F113660%2Fef75d3d8291928dc770b3bd45791b842%2FSelection_999(4961).png?generation=1707365306046858&alt=media)\n\nwonder if the oragnizer is keen to ask kaggler to devlope this?\n\n\"We standardized the resolution, addressed imaging variations, and meticulously annotated blood vessels through semi-automatic,\nmanual, and quality control processes, ensuring high-quality 3D segmentation. Furthermore, we developed a zero-shot cortical blood vessel segmentation method named TriSAM, which leverages the powerful segmentation model SAM for 3D\nsegmentation. To lift SAM from 2D segmentation to 3D volume segmentation, TriSAM employs a multi-seed tracking framework, leveraging the reliability of certain image planes for tracking while using others to identify potential turning points. This approach, consisting of Tri-Plane selection, SAM-based tracking, and recursive redirection, effectively achieves long-term 3D blood vessel segmentation\nwithout model training or fine-tuning. \"", "2642268": "![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F113660%2Fa344bc7d99b126bcdfb543ebb0e7c4f3%2FSelection_999(4963).png?generation=1707365581244185&alt=media)" }, "source": "meta" }