{
  "id": 245107,
  "title": "About data augmentation methods",
  "url": "/competitions/siim-covid19-detection/discussion/245107",
  "author_name": "",
  "post_date": "2021-06-09T17:05:49.145798900Z",
  "votes": null,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Hi, I'm an ML beginner.</p>\n<p>The only data augmentation methods I know of are the general ones, including the ideas already posted. I have an idea for a new idea, but I can't find any documentation on it, nor can I implement it on my own.<br>\nTherefore, I would like to ask you who have a lot of knowledge to share information and advice with me.</p>\n<p>My idea is to remove the bony parts from the radiographs. Will this be an application of GAN or other technology that can remove shadows?　If this idea is implementable, I think it's worth a try. What do you guys think?</p>",
  "messages": [
    {
      "id": "1342717",
      "postDate": "06/09/2021 17:05:49",
      "content": "<p>Hi, I'm an ML beginner.</p>\n<p>The only data augmentation methods I know of are the general ones, including the ideas already posted. I have an idea for a new idea, but I can't find any documentation on it, nor can I implement it on my own.<br>\nTherefore, I would like to ask you who have a lot of knowledge to share information and advice with me.</p>\n<p>My idea is to remove the bony parts from the radiographs. Will this be an application of GAN or other technology that can remove shadows?　If this idea is implementable, I think it's worth a try. What do you guys think?</p>",
      "rawMarkdown": "Hi, I'm an ML beginner.\n\nThe only data augmentation methods I know of are the general ones, including the ideas already posted. I have an idea for a new idea, but I can't find any documentation on it, nor can I implement it on my own.\nTherefore, I would like to ask you who have a lot of knowledge to share information and advice with me.\n\nMy idea is to remove the bony parts from the radiographs. Will this be an application of GAN or other technology that can remove shadows?　If this idea is implementable, I think it's worth a try. What do you guys think?",
      "votes": null
    },
    {
      "id": "1342722",
      "postDate": "06/09/2021 17:17:16",
      "content": "<p>I made a notebook yesterday that uses simple masking to remove rib shadows. </p>\n<p><a href=\"https://www.kaggle.com/davidbroberts/rib-suppression-poc\" target=\"_blank\">https://www.kaggle.com/davidbroberts/rib-suppression-poc</a></p>\n<p>It's similar to digital subtraction that is commonly used in angiography. This is just a brief example, but could be tuned with some gradient filters, segmentation and more masks to create something useful.</p>",
      "rawMarkdown": "I made a notebook yesterday that uses simple masking to remove rib shadows. \n\nhttps://www.kaggle.com/davidbroberts/rib-suppression-poc\n\nIt's similar to digital subtraction that is commonly used in angiography. This is just a brief example, but could be tuned with some gradient filters, segmentation and more masks to create something useful.",
      "votes": null
    },
    {
      "id": "1345241",
      "postDate": "06/11/2021 12:10:15",
      "content": "<p>Thank you for sharing this valuable information. I will study with this NOTEBOOK.</p>",
      "rawMarkdown": "Thank you for sharing this valuable information. I will study with this NOTEBOOK.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1342722,
      "author_name": "davidbroberts",
      "author_url": "",
      "post_date": "06/09/2021 17:17:16",
      "content": "<p>I made a notebook yesterday that uses simple masking to remove rib shadows. </p>\n<p><a href=\"https://www.kaggle.com/davidbroberts/rib-suppression-poc\" target=\"_blank\">https://www.kaggle.com/davidbroberts/rib-suppression-poc</a></p>\n<p>It's similar to digital subtraction that is commonly used in angiography. This is just a brief example, but could be tuned with some gradient filters, segmentation and more masks to create something useful.</p>",
      "votes": null,
      "replies": [
        {
          "id": 1345241,
          "author_name": "kotashimomura",
          "author_url": "",
          "post_date": "06/11/2021 12:10:15",
          "content": "<p>Thank you for sharing this valuable information. I will study with this NOTEBOOK.</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1342717": "Hi, I'm an ML beginner.\n\nThe only data augmentation methods I know of are the general ones, including the ideas already posted. I have an idea for a new idea, but I can't find any documentation on it, nor can I implement it on my own.\nTherefore, I would like to ask you who have a lot of knowledge to share information and advice with me.\n\nMy idea is to remove the bony parts from the radiographs. Will this be an application of GAN or other technology that can remove shadows?　If this idea is implementable, I think it's worth a try. What do you guys think?",
    "1342722": "I made a notebook yesterday that uses simple masking to remove rib shadows. \n\nhttps://www.kaggle.com/davidbroberts/rib-suppression-poc\n\nIt's similar to digital subtraction that is commonly used in angiography. This is just a brief example, but could be tuned with some gradient filters, segmentation and more masks to create something useful.",
    "1345241": "Thank you for sharing this valuable information. I will study with this NOTEBOOK."
  },
  "source": "meta"
}