{
  "id": 182142,
  "title": "Typical FVC calculation",
  "url": "/competitions/osic-pulmonary-fibrosis-progression/discussion/182142",
  "author_name": "",
  "post_date": "2020-09-11T12:28:59.859834400Z",
  "votes": 3,
  "comment_count": 3,
  "views": 0,
  "content": "<p>Here I am sharing the coefficient table of the typical FVC calculation.<br>\nI am sure many people using typical FVC and it is approximated by <code>age</code>, <code>height</code> and <code>race</code>.<br>\n<img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F5055010%2F4e8c33654d1edbfaf5179e327b8a3e8d%2FScreen%20Shot%202020-09-11%20at%2020.14.38.png?generation=1599827601680356&amp;alt=media\" alt=\"\"><br>\nHere is <strong>equation</strong> of typical FVC calculated in <a href=\"https://dynamicmt.com/nhanesIII.pdf\" target=\"_blank\">NHANES III paper</a>:</p>\n<pre><code>FVC (liters) = b0 + b1 * age + b2 * age * age + b3 * height * height\n</code></pre>\n<p>But coefficients (b0, b1, b2, and b3) differ in terms of the race of the patient.<br>\nI am sharing this file because, however it's already written in the paper above, it was little time-consuming to get a table we need.</p>",
  "messages": [
    {
      "id": "1006626",
      "postDate": "09/11/2020 12:28:59",
      "content": "<p>Here I am sharing the coefficient table of the typical FVC calculation.<br>\nI am sure many people using typical FVC and it is approximated by <code>age</code>, <code>height</code> and <code>race</code>.<br>\n<img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F5055010%2F4e8c33654d1edbfaf5179e327b8a3e8d%2FScreen%20Shot%202020-09-11%20at%2020.14.38.png?generation=1599827601680356&amp;alt=media\" alt=\"\"><br>\nHere is <strong>equation</strong> of typical FVC calculated in <a href=\"https://dynamicmt.com/nhanesIII.pdf\" target=\"_blank\">NHANES III paper</a>:</p>\n<pre><code>FVC (liters) = b0 + b1 * age + b2 * age * age + b3 * height * height\n</code></pre>\n<p>But coefficients (b0, b1, b2, and b3) differ in terms of the race of the patient.<br>\nI am sharing this file because, however it's already written in the paper above, it was little time-consuming to get a table we need.</p>",
      "rawMarkdown": "Here I am sharing the coefficient table of the typical FVC calculation.\nI am sure many people using typical FVC and it is approximated by `age`, `height` and `race`.\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F5055010%2F4e8c33654d1edbfaf5179e327b8a3e8d%2FScreen%20Shot%202020-09-11%20at%2020.14.38.png?generation=1599827601680356&alt=media)\nHere is **equation** of typical FVC calculated in [NHANES III paper](https://dynamicmt.com/nhanesIII.pdf):\n```\nFVC (liters) = b0 + b1 * age + b2 * age * age + b3 * height * height\n```\nBut coefficients (b0, b1, b2, and b3) differ in terms of the race of the patient.\nI am sharing this file because, however it's already written in the paper above, it was little time-consuming to get a table we need.",
      "votes": null
    },
    {
      "id": "1006916",
      "postDate": "09/11/2020 16:34:33",
      "content": "<p>Hi! <a href=\"https://www.kaggle.com/bayartsogtya\" target=\"_blank\">@bayartsogtya</a>, <br>\nThanks for the information and the attached CSV, I been trying to use this information but it doesn't improve too much my model performance, How have you use this information in your models?</p>",
      "rawMarkdown": "Hi! @bayartsogtya, \nThanks for the information and the attached CSV, I been trying to use this information but it doesn't improve too much my model performance, How have you use this information in your models?",
      "votes": null
    },
    {
      "id": "1007614",
      "postDate": "09/12/2020 10:20:53",
      "content": "<p>hi <a href=\"https://www.kaggle.com/cv13j0\" target=\"_blank\">@cv13j0</a> <br>\nIn my case it did not help MLOSS based approaches but did help a little on tree-based models.<br>\nI tried this data on multiple models, it really depends on what you are using.<br>\nAlso I am not sure if I am using this information \"correctly\", so that's why I am sharing it too.</p>",
      "rawMarkdown": "hi @cv13j0 \nIn my case it did not help MLOSS based approaches but did help a little on tree-based models.\nI tried this data on multiple models, it really depends on what you are using.\nAlso I am not sure if I am using this information \"correctly\", so that's why I am sharing it too.",
      "votes": null
    },
    {
      "id": "1010699",
      "postDate": "09/15/2020 02:41:43",
      "content": "<p>Hi <a href=\"https://www.kaggle.com/bayartsogtya\" target=\"_blank\">@bayartsogtya</a>,<br>\nThanks for sharing the CSV. It appears to be useful but I'm not sure how to use these parameters as we don't have any information about race and height for each patient. Do you use mean for the race and set a fixed number (such as the mean height of gender) for height?</p>",
      "rawMarkdown": "Hi @bayartsogtya,\nThanks for sharing the CSV. It appears to be useful but I'm not sure how to use these parameters as we don't have any information about race and height for each patient. Do you use mean for the race and set a fixed number (such as the mean height of gender) for height?",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1006916,
      "author_name": "cv13j0",
      "author_url": "",
      "post_date": "09/11/2020 16:34:33",
      "content": "<p>Hi! <a href=\"https://www.kaggle.com/bayartsogtya\" target=\"_blank\">@bayartsogtya</a>, <br>\nThanks for the information and the attached CSV, I been trying to use this information but it doesn't improve too much my model performance, How have you use this information in your models?</p>",
      "votes": null,
      "replies": [
        {
          "id": 1007614,
          "author_name": "bayartsogtya",
          "author_url": "",
          "post_date": "09/12/2020 10:20:53",
          "content": "<p>hi <a href=\"https://www.kaggle.com/cv13j0\" target=\"_blank\">@cv13j0</a> <br>\nIn my case it did not help MLOSS based approaches but did help a little on tree-based models.<br>\nI tried this data on multiple models, it really depends on what you are using.<br>\nAlso I am not sure if I am using this information \"correctly\", so that's why I am sharing it too.</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1010699,
      "author_name": "riow1983",
      "author_url": "",
      "post_date": "09/15/2020 02:41:43",
      "content": "<p>Hi <a href=\"https://www.kaggle.com/bayartsogtya\" target=\"_blank\">@bayartsogtya</a>,<br>\nThanks for sharing the CSV. It appears to be useful but I'm not sure how to use these parameters as we don't have any information about race and height for each patient. Do you use mean for the race and set a fixed number (such as the mean height of gender) for height?</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1006626": "Here I am sharing the coefficient table of the typical FVC calculation.\nI am sure many people using typical FVC and it is approximated by `age`, `height` and `race`.\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F5055010%2F4e8c33654d1edbfaf5179e327b8a3e8d%2FScreen%20Shot%202020-09-11%20at%2020.14.38.png?generation=1599827601680356&alt=media)\nHere is **equation** of typical FVC calculated in [NHANES III paper](https://dynamicmt.com/nhanesIII.pdf):\n```\nFVC (liters) = b0 + b1 * age + b2 * age * age + b3 * height * height\n```\nBut coefficients (b0, b1, b2, and b3) differ in terms of the race of the patient.\nI am sharing this file because, however it's already written in the paper above, it was little time-consuming to get a table we need.",
    "1006916": "Hi! @bayartsogtya, \nThanks for the information and the attached CSV, I been trying to use this information but it doesn't improve too much my model performance, How have you use this information in your models?",
    "1007614": "hi @cv13j0 \nIn my case it did not help MLOSS based approaches but did help a little on tree-based models.\nI tried this data on multiple models, it really depends on what you are using.\nAlso I am not sure if I am using this information \"correctly\", so that's why I am sharing it too.",
    "1010699": "Hi @bayartsogtya,\nThanks for sharing the CSV. It appears to be useful but I'm not sure how to use these parameters as we don't have any information about race and height for each patient. Do you use mean for the race and set a fixed number (such as the mean height of gender) for height?"
  },
  "source": "meta"
}