{
  "id": 429577,
  "title": "Collect only images with lesions",
  "url": "/competitions/rsna-2023-abdominal-trauma-detection/discussion/429577",
  "author_name": "stpete_ishii",
  "post_date": "2023-08-06T09:05:48.065000",
  "votes": 2,
  "comment_count": 2,
  "views": 0,
  "content": "<p>I want to collect only images with lesions.<br>\nThere is too many slices with no lesions even in patients with some abnormality.<br>\nI want to collect only the images with lesions. How can I do it?</p>",
  "messages": [
    {
      "id": 2377573,
      "postDate": "2023-08-07T08:02:46.200Z",
      "content": "<p>A segmentation image is given for some ids, and the organ is segmented.</p>\n<p>To use this data,</p>\n<ol>\n<li>Accurately match segmentation slices with CT scan slices.</li>\n<li>Create a model that can accurately predict the region of each organ in the CT scan image.</li>\n<li>According to the prediction result, create an image that masks everything other than the target organ.</li>\n<li>Learn the presence or absence of organ lesions for images masked other than the target organ.</li>\n</ol>\n<p>If every step is achieved successfully, good score is expected. </p>",
      "rawMarkdown": "A segmentation image is given for some ids, and the organ is segmented.\n\nTo use this data,\n1. Accurately match segmentation slices with CT scan slices.\n2. Create a model that can accurately predict the region of each organ in the CT scan image.\n3. According to the prediction result, create an image that masks everything other than the target organ.\n4. Learn the presence or absence of organ lesions for images masked other than the target organ.\n\nIf every step is achieved successfully, good score is expected. ",
      "votes": 1
    },
    {
      "id": 2376259,
      "postDate": "2023-08-06T09:05:48.067Z",
      "content": "<p>I want to collect only images with lesions.<br>\nThere is too many slices with no lesions even in patients with some abnormality.<br>\nI want to collect only the images with lesions. How can I do it?</p>",
      "rawMarkdown": "I want to collect only images with lesions.\nThere is too many slices with no lesions even in patients with some abnormality.\nI want to collect only the images with lesions. How can I do it?",
      "votes": 2
    },
    {
      "id": 2376292,
      "postDate": "2023-08-06T09:25:47.753Z",
      "content": "<p>Your problem is the title we discussed in the past few days. We will not address this problem if competition host only provides patient level labels and does not provide labels for each individual image. </p>",
      "rawMarkdown": "Your problem is the title we discussed in the past few days. We will not address this problem if competition host only provides patient level labels and does not provide labels for each individual image. "
    }
  ],
  "comments": [
    {
      "id": 2377573,
      "author_name": "stpete_ishii",
      "author_url": "",
      "post_date": "2023-08-07T08:02:46.200000",
      "content": "<p>A segmentation image is given for some ids, and the organ is segmented.</p>\n<p>To use this data,</p>\n<ol>\n<li>Accurately match segmentation slices with CT scan slices.</li>\n<li>Create a model that can accurately predict the region of each organ in the CT scan image.</li>\n<li>According to the prediction result, create an image that masks everything other than the target organ.</li>\n<li>Learn the presence or absence of organ lesions for images masked other than the target organ.</li>\n</ol>\n<p>If every step is achieved successfully, good score is expected. </p>",
      "votes": 1,
      "replies": []
    },
    {
      "id": 2376292,
      "author_name": "Vietnamese-NHNAM",
      "author_url": "",
      "post_date": "2023-08-06T09:25:47.753000",
      "content": "<p>Your problem is the title we discussed in the past few days. We will not address this problem if competition host only provides patient level labels and does not provide labels for each individual image. </p>",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2377573": "A segmentation image is given for some ids, and the organ is segmented.\n\nTo use this data,\n1. Accurately match segmentation slices with CT scan slices.\n2. Create a model that can accurately predict the region of each organ in the CT scan image.\n3. According to the prediction result, create an image that masks everything other than the target organ.\n4. Learn the presence or absence of organ lesions for images masked other than the target organ.\n\nIf every step is achieved successfully, good score is expected. ",
    "2376259": "I want to collect only images with lesions.\nThere is too many slices with no lesions even in patients with some abnormality.\nI want to collect only the images with lesions. How can I do it?",
    "2376292": "Your problem is the title we discussed in the past few days. We will not address this problem if competition host only provides patient level labels and does not provide labels for each individual image. "
  }
}