{
  "id": 97611,
  "title": "One of the Google's Paper for Detection of Diabetic Retinopathy",
  "url": "/competitions/aptos2019-blindness-detection/discussion/97611",
  "author_name": "",
  "post_date": "2019-06-28T01:55:34.827693300Z",
  "votes": 23,
  "comment_count": 2,
  "views": 0,
  "content": "<p>I've read this paper before entitled 'Development and Validation of a Deep Learning Algorithm for Detection of Diabetic Retinopathy in Retinal Fundus Photographs'. In this paper, deep convolutional neural network are trained for detecting diabetic retinopathy. I found this interesting not only in technical aspect but also the application aspect.\n<a href=\"https://jamanetwork.com/journals/jama/fullarticle/2588763\">https://jamanetwork.com/journals/jama/fullarticle/2588763</a></p>",
  "messages": [
    {
      "id": "563178",
      "postDate": "06/28/2019 01:55:34",
      "content": "<p>I've read this paper before entitled 'Development and Validation of a Deep Learning Algorithm for Detection of Diabetic Retinopathy in Retinal Fundus Photographs'. In this paper, deep convolutional neural network are trained for detecting diabetic retinopathy. I found this interesting not only in technical aspect but also the application aspect.\n<a href=\"https://jamanetwork.com/journals/jama/fullarticle/2588763\">https://jamanetwork.com/journals/jama/fullarticle/2588763</a></p>",
      "rawMarkdown": "I've read this paper before entitled 'Development and Validation of a Deep Learning Algorithm for Detection of Diabetic Retinopathy in Retinal Fundus Photographs'. In this paper, deep convolutional neural network are trained for detecting diabetic retinopathy. I found this interesting not only in technical aspect but also the application aspect.\nhttps://jamanetwork.com/journals/jama/fullarticle/2588763",
      "votes": null
    },
    {
      "id": "563202",
      "postDate": "06/28/2019 02:55:36",
      "content": "<p>One point to note is that this is for binary classification only. </p>",
      "rawMarkdown": "One point to note is that this is for binary classification only.",
      "votes": null
    },
    {
      "id": "563827",
      "postDate": "06/28/2019 18:08:48",
      "content": "<p>I am also happy to see this paper because it gives information on how the images were de-identified and other aspects of the ethical oversight. It's not that I think there's much risk of harm to research participants here (though retinal images are used for identification, so this is akin to having pictures of anonymous people's fingertips), but that it will be required if anyone wants to write up and publish their results.</p>",
      "rawMarkdown": "I am also happy to see this paper because it gives information on how the images were de-identified and other aspects of the ethical oversight. It's not that I think there's much risk of harm to research participants here (though retinal images are used for identification, so this is akin to having pictures of anonymous people's fingertips), but that it will be required if anyone wants to write up and publish their results.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 563202,
      "author_name": "tanlikesmath",
      "author_url": "",
      "post_date": "06/28/2019 02:55:36",
      "content": "<p>One point to note is that this is for binary classification only. </p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 563827,
      "author_name": "claurin1",
      "author_url": "",
      "post_date": "06/28/2019 18:08:48",
      "content": "<p>I am also happy to see this paper because it gives information on how the images were de-identified and other aspects of the ethical oversight. It's not that I think there's much risk of harm to research participants here (though retinal images are used for identification, so this is akin to having pictures of anonymous people's fingertips), but that it will be required if anyone wants to write up and publish their results.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "563178": "I've read this paper before entitled 'Development and Validation of a Deep Learning Algorithm for Detection of Diabetic Retinopathy in Retinal Fundus Photographs'. In this paper, deep convolutional neural network are trained for detecting diabetic retinopathy. I found this interesting not only in technical aspect but also the application aspect.\nhttps://jamanetwork.com/journals/jama/fullarticle/2588763",
    "563202": "One point to note is that this is for binary classification only.",
    "563827": "I am also happy to see this paper because it gives information on how the images were de-identified and other aspects of the ethical oversight. It's not that I think there's much risk of harm to research participants here (though retinal images are used for identification, so this is akin to having pictures of anonymous people's fingertips), but that it will be required if anyone wants to write up and publish their results."
  },
  "source": "meta"
}