{
  "id": 174144,
  "title": "Using training CSV data + image pixels data",
  "url": "/competitions/osic-pulmonary-fibrosis-progression/discussion/174144",
  "author_name": "Sovit Ranjan Rath",
  "post_date": "2020-08-12T12:42:16.103000",
  "votes": -1,
  "comment_count": 5,
  "views": 0,
  "content": "<p>I have trained a simple regression model using the training CSV data only. The results are just OK to say the most. But I am also thinking of combining the DICOM Pixels data and training data into one CSV as per the weeks. Has anyone tried it yet or thinking of trying it?</p>",
  "messages": [
    {
      "id": 971759,
      "postDate": "2020-08-15T21:30:33.557Z",
      "content": "<p>Hi,<br>\nI think using DICOM pixels and metadata (csv file) will increase the performance of the model.<br>\nI found this public notebook useful,check it ,it has what you are  looking for.<br>\n<a href=\"https://www.kaggle.com/carlossouza/end-to-end-model-ct-scans-tabular\" target=\"_blank\">https://www.kaggle.com/carlossouza/end-to-end-model-ct-scans-tabular</a><br>\ngood luck.</p>",
      "rawMarkdown": "Hi,\nI think using DICOM pixels and metadata (csv file) will increase the performance of the model.\nI found this public notebook useful,check it ,it has what you are  looking for.\nhttps://www.kaggle.com/carlossouza/end-to-end-model-ct-scans-tabular\ngood luck.",
      "votes": 1,
      "replies": [
        {
          "id": 971843,
          "postDate": "2020-08-16T01:53:33.347Z",
          "content": "<p>Thanks. Will take a look at it.</p>",
          "rawMarkdown": "Thanks. Will take a look at it.",
          "votes": -1
        }
      ]
    },
    {
      "id": 968288,
      "postDate": "2020-08-12T20:41:29.060Z",
      "content": "<p>Don't know if you saw this but people are already using it. <br>\n<a href=\"url\" target=\"_blank\">https://www.kaggle.com/khoongweihao/efficientnets-quantile-regression-inference</a></p>",
      "rawMarkdown": "Don't know if you saw this but people are already using it. \n[https://www.kaggle.com/khoongweihao/efficientnets-quantile-regression-inference](url)",
      "votes": 2,
      "replies": [
        {
          "id": 968382,
          "postDate": "2020-08-13T00:41:30.720Z",
          "content": "<p>Wow! thanks for the link. Will take a look at it.</p>",
          "rawMarkdown": "Wow! thanks for the link. Will take a look at it.",
          "votes": -1
        }
      ]
    },
    {
      "id": 967696,
      "postDate": "2020-08-12T12:42:16.103Z",
      "content": "<p>I have trained a simple regression model using the training CSV data only. The results are just OK to say the most. But I am also thinking of combining the DICOM Pixels data and training data into one CSV as per the weeks. Has anyone tried it yet or thinking of trying it?</p>",
      "rawMarkdown": "I have trained a simple regression model using the training CSV data only. The results are just OK to say the most. But I am also thinking of combining the DICOM Pixels data and training data into one CSV as per the weeks. Has anyone tried it yet or thinking of trying it?",
      "votes": -1
    },
    {
      "id": 974413,
      "postDate": "2020-08-18T00:14:08.350Z",
      "rawMarkdown": "",
      "isDeleted": true
    }
  ],
  "comments": [
    {
      "id": 971759,
      "author_name": "Azzedine",
      "author_url": "",
      "post_date": "2020-08-15T21:30:33.557000",
      "content": "<p>Hi,<br>\nI think using DICOM pixels and metadata (csv file) will increase the performance of the model.<br>\nI found this public notebook useful,check it ,it has what you are  looking for.<br>\n<a href=\"https://www.kaggle.com/carlossouza/end-to-end-model-ct-scans-tabular\" target=\"_blank\">https://www.kaggle.com/carlossouza/end-to-end-model-ct-scans-tabular</a><br>\ngood luck.</p>",
      "votes": 1,
      "replies": [
        {
          "id": 971843,
          "author_name": "Sovit Ranjan Rath",
          "author_url": "",
          "post_date": "2020-08-16T01:53:33.347000",
          "content": "<p>Thanks. Will take a look at it.</p>",
          "votes": -1,
          "replies": []
        }
      ]
    },
    {
      "id": 968288,
      "author_name": "johnny",
      "author_url": "",
      "post_date": "2020-08-12T20:41:29.060000",
      "content": "<p>Don't know if you saw this but people are already using it. <br>\n<a href=\"url\" target=\"_blank\">https://www.kaggle.com/khoongweihao/efficientnets-quantile-regression-inference</a></p>",
      "votes": 2,
      "replies": [
        {
          "id": 968382,
          "author_name": "Sovit Ranjan Rath",
          "author_url": "",
          "post_date": "2020-08-13T00:41:30.720000",
          "content": "<p>Wow! thanks for the link. Will take a look at it.</p>",
          "votes": -1,
          "replies": []
        }
      ]
    },
    {
      "id": 974413,
      "author_name": "",
      "author_url": "",
      "post_date": "2020-08-18T00:14:08.350000",
      "content": "",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "971759": "Hi,\nI think using DICOM pixels and metadata (csv file) will increase the performance of the model.\nI found this public notebook useful,check it ,it has what you are  looking for.\nhttps://www.kaggle.com/carlossouza/end-to-end-model-ct-scans-tabular\ngood luck.",
    "968288": "Don't know if you saw this but people are already using it. \n[https://www.kaggle.com/khoongweihao/efficientnets-quantile-regression-inference](url)",
    "967696": "I have trained a simple regression model using the training CSV data only. The results are just OK to say the most. But I am also thinking of combining the DICOM Pixels data and training data into one CSV as per the weeks. Has anyone tried it yet or thinking of trying it?",
    "974413": ""
  }
}