{
  "id": 352545,
  "title": "Sagittal view for all patients",
  "url": "/competitions/rsna-2022-cervical-spine-fracture-detection/discussion/352545",
  "author_name": "Andreas Renz",
  "post_date": "2022-09-14T19:14:01.595000",
  "votes": 9,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Hi everyone </p>\n<p>I created a notebook (<a href=\"https://www.kaggle.com/code/andir16/rsna-2022-sagittal-view\" target=\"_blank\">RSNA 2022 - Sagittal view</a>) to get the sagittal view (side view on the spine) for all patients (including the ones without segmentation data) by stacking the dcm files and creating a nii file per patient. </p>\n<p>This is the first result for patient 1.2.826.0.1.3680043.10449.<br>\n<img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F6439577%2F80ef7a22d74df0dc086aec9d9c2648ce%2Fsample_image.png?generation=1663182478405491&amp;alt=media\" alt=\"\"></p>\n<p>Feedback, input, and other ideas are very much appreciated. I hope this can help with identifying the vertebrae. </p>",
  "messages": [
    {
      "id": 1939543,
      "postDate": "2022-09-14T19:14:01.597Z",
      "content": "<p>Hi everyone </p>\n<p>I created a notebook (<a href=\"https://www.kaggle.com/code/andir16/rsna-2022-sagittal-view\" target=\"_blank\">RSNA 2022 - Sagittal view</a>) to get the sagittal view (side view on the spine) for all patients (including the ones without segmentation data) by stacking the dcm files and creating a nii file per patient. </p>\n<p>This is the first result for patient 1.2.826.0.1.3680043.10449.<br>\n<img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F6439577%2F80ef7a22d74df0dc086aec9d9c2648ce%2Fsample_image.png?generation=1663182478405491&amp;alt=media\" alt=\"\"></p>\n<p>Feedback, input, and other ideas are very much appreciated. I hope this can help with identifying the vertebrae. </p>",
      "rawMarkdown": "Hi everyone \n\nI created a notebook ([RSNA 2022 - Sagittal view](https://www.kaggle.com/code/andir16/rsna-2022-sagittal-view)) to get the sagittal view (side view on the spine) for all patients (including the ones without segmentation data) by stacking the dcm files and creating a nii file per patient. \n\nThis is the first result for patient 1.2.826.0.1.3680043.10449.\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F6439577%2F80ef7a22d74df0dc086aec9d9c2648ce%2Fsample_image.png?generation=1663182478405491&alt=media)\n\nFeedback, input, and other ideas are very much appreciated. I hope this can help with identifying the vertebrae. ",
      "votes": 9
    },
    {
      "id": 1956430,
      "postDate": "2022-09-26T13:02:36.750Z",
      "content": "<p>Interesting.  You could try to train an object detection network on these images, using the segmentation data as a training set (after a similar conversion to sagittal view).  Maybe some breaks are more detectable on this view.</p>",
      "rawMarkdown": "Interesting.  You could try to train an object detection network on these images, using the segmentation data as a training set (after a similar conversion to sagittal view).  Maybe some breaks are more detectable on this view.",
      "votes": 1,
      "replies": [
        {
          "id": 1957076,
          "postDate": "2022-09-26T18:25:40.557Z",
          "content": "<p>That is an exciting thought. Thank you a lot for your input. I didn't think of using it for detection as well. :) But maybe it can reveal some things better.</p>",
          "rawMarkdown": "That is an exciting thought. Thank you a lot for your input. I didn't think of using it for detection as well. :) But maybe it can reveal some things better."
        }
      ]
    }
  ],
  "comments": [
    {
      "id": 1956430,
      "author_name": "SolverWorld",
      "author_url": "",
      "post_date": "2022-09-26T13:02:36.750000",
      "content": "<p>Interesting.  You could try to train an object detection network on these images, using the segmentation data as a training set (after a similar conversion to sagittal view).  Maybe some breaks are more detectable on this view.</p>",
      "votes": 1,
      "replies": [
        {
          "id": 1957076,
          "author_name": "Andreas Renz",
          "author_url": "",
          "post_date": "2022-09-26T18:25:40.557000",
          "content": "<p>That is an exciting thought. Thank you a lot for your input. I didn't think of using it for detection as well. :) But maybe it can reveal some things better.</p>",
          "votes": 0,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1939543": "Hi everyone \n\nI created a notebook ([RSNA 2022 - Sagittal view](https://www.kaggle.com/code/andir16/rsna-2022-sagittal-view)) to get the sagittal view (side view on the spine) for all patients (including the ones without segmentation data) by stacking the dcm files and creating a nii file per patient. \n\nThis is the first result for patient 1.2.826.0.1.3680043.10449.\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F6439577%2F80ef7a22d74df0dc086aec9d9c2648ce%2Fsample_image.png?generation=1663182478405491&alt=media)\n\nFeedback, input, and other ideas are very much appreciated. I hope this can help with identifying the vertebrae. ",
    "1956430": "Interesting.  You could try to train an object detection network on these images, using the segmentation data as a training set (after a similar conversion to sagittal view).  Maybe some breaks are more detectable on this view."
  }
}