{ "id": 412683, "title": " 📊Interactive visualization of tile annotations", "url": "/competitions/hubmap-hacking-the-human-vasculature/discussion/412683", "author_name": "", "post_date": "2023-05-24T19:54:42.330736Z", "votes": 11, "comment_count": 4, "views": 0, "content": "

I made a notebook in which I provided EDA for the competition's data and created interactive visualization of tile annotations.

\n

Short description:

\n\n
\n

Here is an example visualization of a tile annotations (in the notebook):

\n

\"\"

", "messages": [ { "id": "2272855", "postDate": "05/24/2023 19:54:42", "content": "

I made a notebook in which I provided EDA for the competition's data and created interactive visualization of tile annotations.

\n

Short description:

\n\n
\n

Here is an example visualization of a tile annotations (in the notebook):

\n

\"\"

", "rawMarkdown": "I made a [notebook](https://www.kaggle.com/code/leonidkulyk/eda-hubmap-hhv-interactive-annotations) in which I provided EDA for the competition's data and created interactive visualization of tile annotations.\n\nShort description:\n* Goal: The goal of the competition is to develop a model that can segment instances of microvascular structures, such as capillaries, arterioles, and venules, in 2D PAS-stained histology images from healthy human kidney tissue slides.\n\n* Importance: Automating the segmentation of microvasculature structures will help improve researchers' understanding of how blood vessels are arranged in human tissues. This knowledge is crucial for studying the interaction, organization, and specialization of cells in the body.\n\n***\n\nHere is an example visualization of a tile annotations (in the [notebook](https://www.kaggle.com/code/leonidkulyk/eda-hubmap-hhv-interactive-annotations)):\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F4158783%2F61a90a1661881b1607531bcb443b488f%2Fimage_2023-05-24_22-53-07.png?generation=1684958049039597&alt=media)", "votes": null }, { "id": "2272935", "postDate": "05/24/2023 21:02:58", "content": "

@leonidkulyk wow! I love your interactive visualizations😍

", "rawMarkdown": "leonidkulyk wow! I love your interactive visualizations😍", "votes": null }, { "id": "2274262", "postDate": "05/25/2023 18:28:49", "content": "

Hi;

\n

could you please clarify for me. I understand that usually for medical image segmentation, there is a mask (ground truth), but I didn't see (GT) in the dataset for training.
\ndoes this mean for example that we need to pick a mask?

\n

Thanks!

", "rawMarkdown": "Hi;\n\ncould you please clarify for me. I understand that usually for medical image segmentation, there is a mask (ground truth), but I didn't see (GT) in the dataset for training. \ndoes this mean for example that we need to pick a mask?\n\nThanks!", "votes": null }, { "id": "2274280", "postDate": "05/25/2023 18:59:16", "content": "

Hello, the ground truth segmentations are given in the polygons.jsonl file.

", "rawMarkdown": "Hello, the ground truth segmentations are given in the polygons.jsonl file.", "votes": null }, { "id": "2274296", "postDate": "05/25/2023 19:20:56", "content": "

Thank you Yashvardhan!

", "rawMarkdown": "Thank you Yashvardhan!", "votes": null } ], "comments": [ { "id": 2272935, "author_name": "ihelon", "author_url": "", "post_date": "05/24/2023 21:02:58", "content": "

@leonidkulyk wow! I love your interactive visualizations😍

", "votes": null, "replies": [] }, { "id": 2274262, "author_name": "alqurri", "author_url": "", "post_date": "05/25/2023 18:28:49", "content": "

Hi;

\n

could you please clarify for me. I understand that usually for medical image segmentation, there is a mask (ground truth), but I didn't see (GT) in the dataset for training.
\ndoes this mean for example that we need to pick a mask?

\n

Thanks!

", "votes": null, "replies": [ { "id": 2274280, "author_name": "yashvrdnjain", "author_url": "", "post_date": "05/25/2023 18:59:16", "content": "

Hello, the ground truth segmentations are given in the polygons.jsonl file.

", "votes": null, "replies": [ { "id": 2274296, "author_name": "alqurri", "author_url": "", "post_date": "05/25/2023 19:20:56", "content": "

Thank you Yashvardhan!

", "votes": null, "replies": [] } ] } ] } ], "raw_markdown_by_id": { "2272855": "I made a [notebook](https://www.kaggle.com/code/leonidkulyk/eda-hubmap-hhv-interactive-annotations) in which I provided EDA for the competition's data and created interactive visualization of tile annotations.\n\nShort description:\n* Goal: The goal of the competition is to develop a model that can segment instances of microvascular structures, such as capillaries, arterioles, and venules, in 2D PAS-stained histology images from healthy human kidney tissue slides.\n\n* Importance: Automating the segmentation of microvasculature structures will help improve researchers' understanding of how blood vessels are arranged in human tissues. This knowledge is crucial for studying the interaction, organization, and specialization of cells in the body.\n\n***\n\nHere is an example visualization of a tile annotations (in the [notebook](https://www.kaggle.com/code/leonidkulyk/eda-hubmap-hhv-interactive-annotations)):\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F4158783%2F61a90a1661881b1607531bcb443b488f%2Fimage_2023-05-24_22-53-07.png?generation=1684958049039597&alt=media)", "2272935": "leonidkulyk wow! I love your interactive visualizations😍", "2274262": "Hi;\n\ncould you please clarify for me. I understand that usually for medical image segmentation, there is a mask (ground truth), but I didn't see (GT) in the dataset for training. \ndoes this mean for example that we need to pick a mask?\n\nThanks!", "2274280": "Hello, the ground truth segmentations are given in the polygons.jsonl file.", "2274296": "Thank you Yashvardhan!" }, "source": "meta" }