{
  "id": 14825,
  "title": "Using expanding circles to detect blood and bulges in veins",
  "url": "/competitions/diabetic-retinopathy-detection/discussion/14825",
  "author_name": "",
  "post_date": "2015-06-21T14:36:36.380Z",
  "votes": null,
  "comment_count": 1,
  "views": 909,
  "content": "<p>How to determine if retinopathy is present in an image.</p>\n<p>Take a circle at each point on the image and see if the average intensity in the circle changes as the circle has its diameter increased. When the intensity changes significantly then note the diameter.</p>\n<p>From sample images of disease the average diameter of a bulging vein will be known and will correspond to some circled in the test image if disease is present. There can also be many larger circles where blood has escaped into the retina.</p>",
  "messages": [
    {
      "id": "82505",
      "postDate": "06/21/2015 14:36:36",
      "content": "<p>How to determine if retinopathy is present in an image.</p>\n<p>Take a circle at each point on the image and see if the average intensity in the circle changes as the circle has its diameter increased. When the intensity changes significantly then note the diameter.</p>\n<p>From sample images of disease the average diameter of a bulging vein will be known and will correspond to some circled in the test image if disease is present. There can also be many larger circles where blood has escaped into the retina.</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "82555",
      "postDate": "06/22/2015 15:30:35",
      "content": "<p>I think that yes it would work, however would be extremely slow given the large number of images that have to be processed.</p>",
      "rawMarkdown": "",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 82555,
      "author_name": "antiquark",
      "author_url": "",
      "post_date": "06/22/2015 15:30:35",
      "content": "<p>I think that yes it would work, however would be extremely slow given the large number of images that have to be processed.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "82505": "",
    "82555": ""
  },
  "source": "meta"
}