{
  "id": 163611,
  "title": "Cuts?",
  "url": "/competitions/siim-isic-melanoma-classification/discussion/163611",
  "author_name": "Tord Malmgren",
  "post_date": "2020-07-02T18:09:11.886000",
  "votes": 1,
  "comment_count": 10,
  "views": 0,
  "content": "<p>Are these \"cuts\" sampling of the melanoma?<img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F199197%2Fd9f14ff4e463d012b00beaf03f753e39%2FCuts.PNG?generation=1593713335529260&amp;alt=media\" alt=\"\"></p>\n\n<p>Or some part of a tool? It's visible in more than one image.</p>",
  "messages": [
    {
      "id": 912883,
      "postDate": "2020-07-02T19:31:52.227Z",
      "content": "<p>I think it has something to do with the way the image is taken. If you look at this reference, you will see similar gridlines.</p>\n\n<p><a href=\"https://www.researchgate.net/publication/6495721_Correlation_of_Dermoscopic_Structures_of_Melanocytic_Lesions_to_Reflectance_Confocal_Microscopy/link/00b7d53625868ecd8c000000/download\">https://www.researchgate.net/publication/6495721_Correlation_of_Dermoscopic_Structures_of_Melanocytic_Lesions_to_Reflectance_Confocal_Microscopy/link/00b7d53625868ecd8c000000/download</a></p>",
      "rawMarkdown": "I think it has something to do with the way the image is taken. If you look at this reference, you will see similar gridlines.\n\nhttps://www.researchgate.net/publication/6495721_Correlation_of_Dermoscopic_Structures_of_Melanocytic_Lesions_to_Reflectance_Confocal_Microscopy/link/00b7d53625868ecd8c000000/download",
      "votes": 1,
      "replies": [
        {
          "id": 914874,
          "postDate": "2020-07-04T10:02:52.473Z",
          "content": "<p>Yes! Looks exactly like the mosaic of Reflectance Confocal Microscopy (RCM). Thanks!</p>",
          "rawMarkdown": "Yes! Looks exactly like the mosaic of Reflectance Confocal Microscopy (RCM). Thanks!"
        }
      ]
    },
    {
      "id": 912789,
      "postDate": "2020-07-02T18:13:28.733Z",
      "content": "<p>Are you referring to the whitish lines, which are probably hair overlapping the lesion?</p>",
      "rawMarkdown": "Are you referring to the whitish lines, which are probably hair overlapping the lesion?",
      "votes": 1,
      "replies": [
        {
          "id": 912794,
          "postDate": "2020-07-02T18:17:42.787Z",
          "content": "<p>No, the dark equidistantial vertical lines (in the right figure, which is my transformed image)</p>",
          "rawMarkdown": "No, the dark equidistantial vertical lines (in the right figure, which is my transformed image)"
        }
      ]
    },
    {
      "id": 912818,
      "postDate": "2020-07-02T18:44:40.007Z",
      "content": "<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F199197%2Ff63d5e45919599c4faeffc084c2e200a%2FKaggle.png?generation=1593715447000748&amp;alt=media\" alt=\"\"></p>\n\n<p>With identifier... I just wonder if I'm implementing a bug...</p>",
      "rawMarkdown": "![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F199197%2Ff63d5e45919599c4faeffc084c2e200a%2FKaggle.png?generation=1593715447000748&amp;alt=media)\n\nWith identifier... I just wonder if I'm implementing a bug...",
      "replies": [
        {
          "id": 913571,
          "postDate": "2020-07-03T09:32:52.410Z",
          "content": "<p>If you look closely at the RGB images you can even see it there (\"green-ish\"). It is, in my experience, an image artifact from one common digital video-dermatoscopic imaging device. I don't know the root cause (PAL Video encoding / interpolation / ...) though.</p>",
          "rawMarkdown": "If you look closely at the RGB images you can even see it there (\"green-ish\"). It is, in my experience, an image artifact from one common digital video-dermatoscopic imaging device. I don't know the root cause (PAL Video encoding / interpolation / ...) though.",
          "votes": 1
        }
      ]
    },
    {
      "id": 912784,
      "postDate": "2020-07-02T18:09:11.887Z",
      "content": "<p>Are these \"cuts\" sampling of the melanoma?<img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F199197%2Fd9f14ff4e463d012b00beaf03f753e39%2FCuts.PNG?generation=1593713335529260&amp;alt=media\" alt=\"\"></p>\n\n<p>Or some part of a tool? It's visible in more than one image.</p>",
      "rawMarkdown": "Are these \"cuts\" sampling of the melanoma?![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F199197%2Fd9f14ff4e463d012b00beaf03f753e39%2FCuts.PNG?generation=1593713335529260&amp;alt=media)\n\nOr some part of a tool? It's visible in more than one image.\n"
    },
    {
      "id": 913309,
      "postDate": "2020-07-03T06:10:13.263Z",
      "rawMarkdown": "",
      "votes": 1,
      "isDeleted": true,
      "replies": [
        {
          "id": 914886,
          "postDate": "2020-07-04T10:09:44.993Z",
          "content": "<p>For 0091996, I used\ngray = .2 r - .00171 r^2 + .00444 rg - .00189 rb - .435 g -.00255 g^2 + .00278 gb + .134 b - .000469 b^2</p>",
          "rawMarkdown": "For 0091996, I used\ngray = .2 r - .00171 r^2 + .00444 rg - .00189 rb - .435 g -.00255 g^2 + .00278 gb + .134 b - .000469 b^2",
          "votes": 1
        },
        {
          "id": 914927,
          "postDate": "2020-07-04T11:03:33.860Z",
          "rawMarkdown": "",
          "isDeleted": true
        },
        {
          "id": 915177,
          "postDate": "2020-07-04T14:18:55.220Z",
          "content": "<p>I was trying to preprocess and assumed the interesting stuff is in the center, so using three bands to discriminate between \"inside bounding box\" and \"outside\" should yield a grayscale image for convolution analysis. I am starting to re-think this now...</p>",
          "rawMarkdown": "I was trying to preprocess and assumed the interesting stuff is in the center, so using three bands to discriminate between \"inside bounding box\" and \"outside\" should yield a grayscale image for convolution analysis. I am starting to re-think this now...",
          "votes": 1
        }
      ]
    }
  ],
  "comments": [
    {
      "id": 912883,
      "author_name": "quadcore/Richard Epstein",
      "author_url": "",
      "post_date": "2020-07-02T19:31:52.227000",
      "content": "<p>I think it has something to do with the way the image is taken. If you look at this reference, you will see similar gridlines.</p>\n\n<p><a href=\"https://www.researchgate.net/publication/6495721_Correlation_of_Dermoscopic_Structures_of_Melanocytic_Lesions_to_Reflectance_Confocal_Microscopy/link/00b7d53625868ecd8c000000/download\">https://www.researchgate.net/publication/6495721_Correlation_of_Dermoscopic_Structures_of_Melanocytic_Lesions_to_Reflectance_Confocal_Microscopy/link/00b7d53625868ecd8c000000/download</a></p>",
      "votes": 1,
      "replies": [
        {
          "id": 914874,
          "author_name": "Tord Malmgren",
          "author_url": "",
          "post_date": "2020-07-04T10:02:52.473000",
          "content": "<p>Yes! Looks exactly like the mosaic of Reflectance Confocal Microscopy (RCM). Thanks!</p>",
          "votes": 0,
          "replies": []
        }
      ]
    },
    {
      "id": 912789,
      "author_name": "Sirish Somanchi",
      "author_url": "",
      "post_date": "2020-07-02T18:13:28.733000",
      "content": "<p>Are you referring to the whitish lines, which are probably hair overlapping the lesion?</p>",
      "votes": 1,
      "replies": [
        {
          "id": 912794,
          "author_name": "Tord Malmgren",
          "author_url": "",
          "post_date": "2020-07-02T18:17:42.787000",
          "content": "<p>No, the dark equidistantial vertical lines (in the right figure, which is my transformed image)</p>",
          "votes": 0,
          "replies": []
        }
      ]
    },
    {
      "id": 912818,
      "author_name": "Tord Malmgren",
      "author_url": "",
      "post_date": "2020-07-02T18:44:40.007000",
      "content": "<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F199197%2Ff63d5e45919599c4faeffc084c2e200a%2FKaggle.png?generation=1593715447000748&amp;alt=media\" alt=\"\"></p>\n\n<p>With identifier... I just wonder if I'm implementing a bug...</p>",
      "votes": 0,
      "replies": [
        {
          "id": 913571,
          "author_name": "chdlr",
          "author_url": "",
          "post_date": "2020-07-03T09:32:52.410000",
          "content": "<p>If you look closely at the RGB images you can even see it there (\"green-ish\"). It is, in my experience, an image artifact from one common digital video-dermatoscopic imaging device. I don't know the root cause (PAL Video encoding / interpolation / ...) though.</p>",
          "votes": 1,
          "replies": []
        }
      ]
    },
    {
      "id": 913309,
      "author_name": "",
      "author_url": "",
      "post_date": "2020-07-03T06:10:13.263000",
      "content": "",
      "votes": 1,
      "replies": [
        {
          "id": 914886,
          "author_name": "Tord Malmgren",
          "author_url": "",
          "post_date": "2020-07-04T10:09:44.993000",
          "content": "<p>For 0091996, I used\ngray = .2 r - .00171 r^2 + .00444 rg - .00189 rb - .435 g -.00255 g^2 + .00278 gb + .134 b - .000469 b^2</p>",
          "votes": 1,
          "replies": []
        },
        {
          "id": 914927,
          "author_name": "",
          "author_url": "",
          "post_date": "2020-07-04T11:03:33.860000",
          "content": "",
          "votes": 0,
          "replies": []
        },
        {
          "id": 915177,
          "author_name": "Tord Malmgren",
          "author_url": "",
          "post_date": "2020-07-04T14:18:55.220000",
          "content": "<p>I was trying to preprocess and assumed the interesting stuff is in the center, so using three bands to discriminate between \"inside bounding box\" and \"outside\" should yield a grayscale image for convolution analysis. I am starting to re-think this now...</p>",
          "votes": 1,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "912883": "I think it has something to do with the way the image is taken. If you look at this reference, you will see similar gridlines.\n\nhttps://www.researchgate.net/publication/6495721_Correlation_of_Dermoscopic_Structures_of_Melanocytic_Lesions_to_Reflectance_Confocal_Microscopy/link/00b7d53625868ecd8c000000/download",
    "912789": "Are you referring to the whitish lines, which are probably hair overlapping the lesion?",
    "912818": "![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F199197%2Ff63d5e45919599c4faeffc084c2e200a%2FKaggle.png?generation=1593715447000748&amp;alt=media)\n\nWith identifier... I just wonder if I'm implementing a bug...",
    "912784": "Are these \"cuts\" sampling of the melanoma?![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F199197%2Fd9f14ff4e463d012b00beaf03f753e39%2FCuts.PNG?generation=1593713335529260&amp;alt=media)\n\nOr some part of a tool? It's visible in more than one image.\n",
    "913309": ""
  }
}