{
  "id": 253020,
  "title": "Notebook for determining MR image planes",
  "url": "/competitions/rsna-miccai-brain-tumor-radiogenomic-classification/discussion/253020",
  "author_name": "",
  "post_date": "2021-07-14T17:25:25.203625100Z",
  "votes": 10,
  "comment_count": 3,
  "views": 0,
  "content": "<p>I noticed that not all the series were in the same plane (relative to the patient's body) across studies. For instance, study A's FLAIR sequence is in the Axial plane, while study B's FLAIR sequence is coronal.</p>\n<p>It is necessary to determine the image plane (axial, sagittal, coronal) from the DICOM tag (0x0020,0x0037 = Image Position Patient).</p>\n<p>So, I made a notebook to explain it.</p>\n<p><a href=\"https://www.kaggle.com/davidbroberts/determining-mr-image-planes\" target=\"_blank\">https://www.kaggle.com/davidbroberts/determining-mr-image-planes</a></p>",
  "messages": [
    {
      "id": "1388161",
      "postDate": "07/14/2021 17:25:25",
      "content": "<p>I noticed that not all the series were in the same plane (relative to the patient's body) across studies. For instance, study A's FLAIR sequence is in the Axial plane, while study B's FLAIR sequence is coronal.</p>\n<p>It is necessary to determine the image plane (axial, sagittal, coronal) from the DICOM tag (0x0020,0x0037 = Image Position Patient).</p>\n<p>So, I made a notebook to explain it.</p>\n<p><a href=\"https://www.kaggle.com/davidbroberts/determining-mr-image-planes\" target=\"_blank\">https://www.kaggle.com/davidbroberts/determining-mr-image-planes</a></p>",
      "rawMarkdown": "I noticed that not all the series were in the same plane (relative to the patient's body) across studies. For instance, study A's FLAIR sequence is in the Axial plane, while study B's FLAIR sequence is coronal.\n\nIt is necessary to determine the image plane (axial, sagittal, coronal) from the DICOM tag (0x0020,0x0037 = Image Position Patient).\n\nSo, I made a notebook to explain it.\n\nhttps://www.kaggle.com/davidbroberts/determining-mr-image-planes",
      "votes": null
    },
    {
      "id": "1388397",
      "postDate": "07/14/2021 22:40:00",
      "content": "<p>Thank you!!</p>",
      "rawMarkdown": "Thank you!!",
      "votes": null
    },
    {
      "id": "1543618",
      "postDate": "10/13/2021 17:26:20",
      "content": "<p>how will your model improve if you decide to utilize mri planes (coronal, sagittal , axial) from dicom metadata , how would you even use in your model?</p>",
      "rawMarkdown": "how will your model improve if you decide to utilize mri planes (coronal, sagittal , axial) from dicom metadata , how would you even use in your model?",
      "votes": null
    },
    {
      "id": "1543734",
      "postDate": "10/13/2021 18:57:24",
      "content": "<p>All images are in one of the three planes. It doesn't make sense to build a CNN with some images that are axial and some that are coronal .. The idea is to make sure all the images are in the same plane before training. Otherwise, your training data will be a mix of planes.</p>",
      "rawMarkdown": "All images are in one of the three planes. It doesn't make sense to build a CNN with some images that are axial and some that are coronal .. The idea is to make sure all the images are in the same plane before training. Otherwise, your training data will be a mix of planes.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1388397,
      "author_name": "elenaeb",
      "author_url": "",
      "post_date": "07/14/2021 22:40:00",
      "content": "<p>Thank you!!</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1543618,
      "author_name": "aniarya",
      "author_url": "",
      "post_date": "10/13/2021 17:26:20",
      "content": "<p>how will your model improve if you decide to utilize mri planes (coronal, sagittal , axial) from dicom metadata , how would you even use in your model?</p>",
      "votes": null,
      "replies": [
        {
          "id": 1543734,
          "author_name": "davidbroberts",
          "author_url": "",
          "post_date": "10/13/2021 18:57:24",
          "content": "<p>All images are in one of the three planes. It doesn't make sense to build a CNN with some images that are axial and some that are coronal .. The idea is to make sure all the images are in the same plane before training. Otherwise, your training data will be a mix of planes.</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1388161": "I noticed that not all the series were in the same plane (relative to the patient's body) across studies. For instance, study A's FLAIR sequence is in the Axial plane, while study B's FLAIR sequence is coronal.\n\nIt is necessary to determine the image plane (axial, sagittal, coronal) from the DICOM tag (0x0020,0x0037 = Image Position Patient).\n\nSo, I made a notebook to explain it.\n\nhttps://www.kaggle.com/davidbroberts/determining-mr-image-planes",
    "1388397": "Thank you!!",
    "1543618": "how will your model improve if you decide to utilize mri planes (coronal, sagittal , axial) from dicom metadata , how would you even use in your model?",
    "1543734": "All images are in one of the three planes. It doesn't make sense to build a CNN with some images that are axial and some that are coronal .. The idea is to make sure all the images are in the same plane before training. Otherwise, your training data will be a mix of planes."
  },
  "source": "meta"
}