{
  "id": 220079,
  "title": "Abdominal Xrays in Dataset",
  "url": "/competitions/ranzcr-clip-catheter-line-classification/discussion/220079",
  "author_name": "",
  "post_date": "2021-02-17T10:16:34.911069600Z",
  "votes": 4,
  "comment_count": 1,
  "views": 0,
  "content": "<p>In going through images with the label 'NGT - Abnormal', I've noticed a few abdominal X-rays, such as</p>\n<p>StudyInstanceUID: 1-2-826-0-1-3680043-8-498-74541722970738045639189285999486415964<br>\n<img src=\"https://i.ibb.co/WvPyMxp/1-2-826-0-1-3680043-8-498-74541722970738045639189285999486415964.jpg\" alt=\"\"></p>\n<p>StudyInstanceUID: 1-2-826-0-1-3680043-8-498-66613400234207033453452780199854965138<br>\n<img src=\"https://i.ibb.co/wzdsJFt/1-2-826-0-1-3680043-8-498-66613400234207033453452780199854965138.jpg\" alt=\"\"></p>\n<p>Out of the 279 images in that class, I counted five abdo XRs. Including this one which is a cool image of a barium swallow follow through:</p>\n<p>StudyInstanceUID: 1-2-826-0-1-3680043-8-498-30321789485569066950941222990558463472<br>\n<img src=\"https://i.ibb.co/FJb4yxh/1-2-826-0-1-3680043-8-498-30321789485569066950941222990558463472.jpg\" alt=\"\"></p>\n<p>There are 4797 images in the NGT - Normal class, which should have a few abdo XRs too. Interestingly they all look like paediatric abdo XRs, which makes me think this dataset could be a useful supplement after all: <a href=\"https://github.com/xinario/PediatricXray100\" target=\"_blank\">https://github.com/xinario/PediatricXray100</a> (thanks Xin Yi)</p>",
  "messages": [
    {
      "id": "1206436",
      "postDate": "02/17/2021 10:16:34",
      "content": "<p>In going through images with the label 'NGT - Abnormal', I've noticed a few abdominal X-rays, such as</p>\n<p>StudyInstanceUID: 1-2-826-0-1-3680043-8-498-74541722970738045639189285999486415964<br>\n<img src=\"https://i.ibb.co/WvPyMxp/1-2-826-0-1-3680043-8-498-74541722970738045639189285999486415964.jpg\" alt=\"\"></p>\n<p>StudyInstanceUID: 1-2-826-0-1-3680043-8-498-66613400234207033453452780199854965138<br>\n<img src=\"https://i.ibb.co/wzdsJFt/1-2-826-0-1-3680043-8-498-66613400234207033453452780199854965138.jpg\" alt=\"\"></p>\n<p>Out of the 279 images in that class, I counted five abdo XRs. Including this one which is a cool image of a barium swallow follow through:</p>\n<p>StudyInstanceUID: 1-2-826-0-1-3680043-8-498-30321789485569066950941222990558463472<br>\n<img src=\"https://i.ibb.co/FJb4yxh/1-2-826-0-1-3680043-8-498-30321789485569066950941222990558463472.jpg\" alt=\"\"></p>\n<p>There are 4797 images in the NGT - Normal class, which should have a few abdo XRs too. Interestingly they all look like paediatric abdo XRs, which makes me think this dataset could be a useful supplement after all: <a href=\"https://github.com/xinario/PediatricXray100\" target=\"_blank\">https://github.com/xinario/PediatricXray100</a> (thanks Xin Yi)</p>",
      "rawMarkdown": "In going through images with the label 'NGT - Abnormal', I've noticed a few abdominal X-rays, such as\n\nStudyInstanceUID: 1-2-826-0-1-3680043-8-498-74541722970738045639189285999486415964\n![](https://i.ibb.co/WvPyMxp/1-2-826-0-1-3680043-8-498-74541722970738045639189285999486415964.jpg)\n\nStudyInstanceUID: 1-2-826-0-1-3680043-8-498-66613400234207033453452780199854965138\n![](https://i.ibb.co/wzdsJFt/1-2-826-0-1-3680043-8-498-66613400234207033453452780199854965138.jpg)\n\nOut of the 279 images in that class, I counted five abdo XRs. Including this one which is a cool image of a barium swallow follow through:\n\nStudyInstanceUID: 1-2-826-0-1-3680043-8-498-30321789485569066950941222990558463472\n![](https://i.ibb.co/FJb4yxh/1-2-826-0-1-3680043-8-498-30321789485569066950941222990558463472.jpg)\n\nThere are 4797 images in the NGT - Normal class, which should have a few abdo XRs too. Interestingly they all look like paediatric abdo XRs, which makes me think this dataset could be a useful supplement after all: https://github.com/xinario/PediatricXray100 (thanks Xin Yi)",
      "votes": null
    },
    {
      "id": "1213152",
      "postDate": "02/21/2021 22:20:08",
      "content": "<p>Having a separate model to identify abdominal view of Xray might be helpful. (since we can't see CVC abd ETT on this view). Also, differentiating between abdominal and Chest view should be easy.<br>\nHas someone tried this approach?  Please do let me know your findings. </p>",
      "rawMarkdown": "Having a separate model to identify abdominal view of Xray might be helpful. (since we can't see CVC abd ETT on this view). Also, differentiating between abdominal and Chest view should be easy.\nHas someone tried this approach?  Please do let me know your findings.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1213152,
      "author_name": "amritpal333",
      "author_url": "",
      "post_date": "02/21/2021 22:20:08",
      "content": "<p>Having a separate model to identify abdominal view of Xray might be helpful. (since we can't see CVC abd ETT on this view). Also, differentiating between abdominal and Chest view should be easy.<br>\nHas someone tried this approach?  Please do let me know your findings. </p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1206436": "In going through images with the label 'NGT - Abnormal', I've noticed a few abdominal X-rays, such as\n\nStudyInstanceUID: 1-2-826-0-1-3680043-8-498-74541722970738045639189285999486415964\n![](https://i.ibb.co/WvPyMxp/1-2-826-0-1-3680043-8-498-74541722970738045639189285999486415964.jpg)\n\nStudyInstanceUID: 1-2-826-0-1-3680043-8-498-66613400234207033453452780199854965138\n![](https://i.ibb.co/wzdsJFt/1-2-826-0-1-3680043-8-498-66613400234207033453452780199854965138.jpg)\n\nOut of the 279 images in that class, I counted five abdo XRs. Including this one which is a cool image of a barium swallow follow through:\n\nStudyInstanceUID: 1-2-826-0-1-3680043-8-498-30321789485569066950941222990558463472\n![](https://i.ibb.co/FJb4yxh/1-2-826-0-1-3680043-8-498-30321789485569066950941222990558463472.jpg)\n\nThere are 4797 images in the NGT - Normal class, which should have a few abdo XRs too. Interestingly they all look like paediatric abdo XRs, which makes me think this dataset could be a useful supplement after all: https://github.com/xinario/PediatricXray100 (thanks Xin Yi)",
    "1213152": "Having a separate model to identify abdominal view of Xray might be helpful. (since we can't see CVC abd ETT on this view). Also, differentiating between abdominal and Chest view should be easy.\nHas someone tried this approach?  Please do let me know your findings."
  },
  "source": "meta"
}