{
  "id": 539436,
  "title": "Converting from dicom to png",
  "url": "/competitions/vinbigdata-chest-xray-abnormalities-detection/discussion/539436",
  "author_name": "Yasser Ali",
  "post_date": "2024-10-08T22:56:26.120000",
  "votes": 0,
  "comment_count": 0,
  "views": 0,
  "content": "<p>I found in this competition it was very helpful converting from dicom to png to reduce the time spent training by a ton. So if training is taking very long, I would recommend converting to png. The size I used was 512, 512 and found that worked pretty well. </p>",
  "messages": [
    {
      "id": 3012325,
      "postDate": "2024-10-08T22:56:26.120Z",
      "content": "<p>I found in this competition it was very helpful converting from dicom to png to reduce the time spent training by a ton. So if training is taking very long, I would recommend converting to png. The size I used was 512, 512 and found that worked pretty well. </p>",
      "rawMarkdown": "I found in this competition it was very helpful converting from dicom to png to reduce the time spent training by a ton. So if training is taking very long, I would recommend converting to png. The size I used was 512, 512 and found that worked pretty well. "
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "3012325": "I found in this competition it was very helpful converting from dicom to png to reduce the time spent training by a ton. So if training is taking very long, I would recommend converting to png. The size I used was 512, 512 and found that worked pretty well. "
  }
}