{
  "id": 452898,
  "title": "Has anyone tried using histolab for this competition?",
  "url": "/competitions/UBC-OCEAN/discussion/452898",
  "author_name": "Cleyton Candeira",
  "post_date": "2023-11-04T01:00:34.299000",
  "votes": 5,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Histolab is a project that allows WSI images to be processed in a practical way, automatically detecting tissue, cleaning images and obtaining mosaics from scores, and can be integrated into a deep learning pipeline (<a href=\"https://histolab.readthedocs.io/en/latest/index.html)\" target=\"_blank\">https://histolab.readthedocs.io/en/latest/index.html)</a>. I tested converting .png files to .tif, suggested by Noli Alonso (<a href=\"https://www.kaggle.com/code/nolialonso/ubc-ocean-png-to-tiff\" target=\"_blank\">https://www.kaggle.com/code/nolialonso/ubc-ocean-png-to-tiff</a>) but I was unfortunate enough to discover that, once the conversion was complete, it was impossible to access the files' property levels. What's more, I tried using to access the data, but I couldn't do it either. </p>\n<p>Histolab can be a perfect project to use in this challenge, but, the .png format and even its conversion does not contain the wealth of pyramid information.</p>\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F7896550%2F8241af6f3f583db2796b807c23c29389%2FCaptura%20de%20tela%20de%202023-11-03%2022-00-04.png?generation=1699059620368631&amp;alt=media\" alt=\"score_example\"></p>",
  "messages": [
    {
      "id": 2511796,
      "postDate": "2023-11-04T01:00:34.300Z",
      "content": "<p>Histolab is a project that allows WSI images to be processed in a practical way, automatically detecting tissue, cleaning images and obtaining mosaics from scores, and can be integrated into a deep learning pipeline (<a href=\"https://histolab.readthedocs.io/en/latest/index.html)\" target=\"_blank\">https://histolab.readthedocs.io/en/latest/index.html)</a>. I tested converting .png files to .tif, suggested by Noli Alonso (<a href=\"https://www.kaggle.com/code/nolialonso/ubc-ocean-png-to-tiff\" target=\"_blank\">https://www.kaggle.com/code/nolialonso/ubc-ocean-png-to-tiff</a>) but I was unfortunate enough to discover that, once the conversion was complete, it was impossible to access the files' property levels. What's more, I tried using to access the data, but I couldn't do it either. </p>\n<p>Histolab can be a perfect project to use in this challenge, but, the .png format and even its conversion does not contain the wealth of pyramid information.</p>\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F7896550%2F8241af6f3f583db2796b807c23c29389%2FCaptura%20de%20tela%20de%202023-11-03%2022-00-04.png?generation=1699059620368631&amp;alt=media\" alt=\"score_example\"></p>",
      "rawMarkdown": "Histolab is a project that allows WSI images to be processed in a practical way, automatically detecting tissue, cleaning images and obtaining mosaics from scores, and can be integrated into a deep learning pipeline (https://histolab.readthedocs.io/en/latest/index.html). I tested converting .png files to .tif, suggested by Noli Alonso (https://www.kaggle.com/code/nolialonso/ubc-ocean-png-to-tiff) but I was unfortunate enough to discover that, once the conversion was complete, it was impossible to access the files' property levels. What's more, I tried using to access the data, but I couldn't do it either. \n\nHistolab can be a perfect project to use in this challenge, but, the .png format and even its conversion does not contain the wealth of pyramid information.\n\n![score_example](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F7896550%2F8241af6f3f583db2796b807c23c29389%2FCaptura%20de%20tela%20de%202023-11-03%2022-00-04.png?generation=1699059620368631&alt=media)\n\n",
      "votes": 5
    },
    {
      "id": 2511889,
      "postDate": "2023-11-04T03:52:11.253Z",
      "content": "<p>I tried it and was not able to make it work. Histolab uses Openslide, which cannot input PNG files, only output them. I tried converting the PNG files to tiff, but then ran into the same properties issues. <br>\nEven if it did work, the time to convert it to tiff consumes way too much of the 12 hours.</p>",
      "rawMarkdown": "I tried it and was not able to make it work. Histolab uses Openslide, which cannot input PNG files, only output them. I tried converting the PNG files to tiff, but then ran into the same properties issues. \nEven if it did work, the time to convert it to tiff consumes way too much of the 12 hours.",
      "votes": 4
    }
  ],
  "comments": [
    {
      "id": 2511889,
      "author_name": "Noli Alonso",
      "author_url": "",
      "post_date": "2023-11-04T03:52:11.253000",
      "content": "<p>I tried it and was not able to make it work. Histolab uses Openslide, which cannot input PNG files, only output them. I tried converting the PNG files to tiff, but then ran into the same properties issues. <br>\nEven if it did work, the time to convert it to tiff consumes way too much of the 12 hours.</p>",
      "votes": 4,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2511796": "Histolab is a project that allows WSI images to be processed in a practical way, automatically detecting tissue, cleaning images and obtaining mosaics from scores, and can be integrated into a deep learning pipeline (https://histolab.readthedocs.io/en/latest/index.html). I tested converting .png files to .tif, suggested by Noli Alonso (https://www.kaggle.com/code/nolialonso/ubc-ocean-png-to-tiff) but I was unfortunate enough to discover that, once the conversion was complete, it was impossible to access the files' property levels. What's more, I tried using to access the data, but I couldn't do it either. \n\nHistolab can be a perfect project to use in this challenge, but, the .png format and even its conversion does not contain the wealth of pyramid information.\n\n![score_example](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F7896550%2F8241af6f3f583db2796b807c23c29389%2FCaptura%20de%20tela%20de%202023-11-03%2022-00-04.png?generation=1699059620368631&alt=media)\n\n",
    "2511889": "I tried it and was not able to make it work. Histolab uses Openslide, which cannot input PNG files, only output them. I tried converting the PNG files to tiff, but then ran into the same properties issues. \nEven if it did work, the time to convert it to tiff consumes way too much of the 12 hours."
  }
}