{
  "id": 609686,
  "title": "How can I obtain the correct modalities？",
  "url": "/competitions/rsna-intracranial-aneurysm-detection/discussion/609686",
  "author_name": "Less",
  "post_date": "2025-09-29T02:36:24.256000",
  "votes": 4,
  "comment_count": 2,
  "views": 0,
  "content": "<p>May I ask a question? The DICOM modalities I read from the test dataset are CT, MR, and MR. Why not CTA, MRA, and MRI? How can I obtain the correct modalities ?</p>",
  "messages": [
    {
      "id": 3295589,
      "postDate": "2025-09-29T07:20:44.837Z",
      "content": "<p>You can't. You can just predict them or use a modality agnostic model.</p>",
      "rawMarkdown": "You can't. You can just predict them or use a modality agnostic model.",
      "votes": 3
    },
    {
      "id": 3295520,
      "postDate": "2025-09-29T02:36:24.257Z",
      "content": "<p>May I ask a question? The DICOM modalities I read from the test dataset are CT, MR, and MR. Why not CTA, MRA, and MRI? How can I obtain the correct modalities ?</p>",
      "rawMarkdown": "May I ask a question? The DICOM modalities I read from the test dataset are CT, MR, and MR. Why not CTA, MRA, and MRI? How can I obtain the correct modalities ?",
      "votes": 4
    },
    {
      "id": 3296794,
      "postDate": "2025-10-01T16:14:56.743Z",
      "content": "<p>MRA is a subset of MRI. In practice there is no straightforward way to distinguish any two given MRI sequences other than looking at the images and the header variables. Given the limited header information available in the test set, I suggest that you use the image data to determine what type of sequence you are dealing with.</p>",
      "rawMarkdown": "MRA is a subset of MRI. In practice there is no straightforward way to distinguish any two given MRI sequences other than looking at the images and the header variables. Given the limited header information available in the test set, I suggest that you use the image data to determine what type of sequence you are dealing with.",
      "votes": 2
    }
  ],
  "comments": [
    {
      "id": 3295589,
      "author_name": "Ángel Jacinto Sánchez Ruiz",
      "author_url": "",
      "post_date": "2025-09-29T07:20:44.837000",
      "content": "<p>You can't. You can just predict them or use a modality agnostic model.</p>",
      "votes": 3,
      "replies": []
    },
    {
      "id": 3296794,
      "author_name": "Evan Calabrese",
      "author_url": "",
      "post_date": "2025-10-01T16:14:56.743000",
      "content": "<p>MRA is a subset of MRI. In practice there is no straightforward way to distinguish any two given MRI sequences other than looking at the images and the header variables. Given the limited header information available in the test set, I suggest that you use the image data to determine what type of sequence you are dealing with.</p>",
      "votes": 2,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "3295589": "You can't. You can just predict them or use a modality agnostic model.",
    "3295520": "May I ask a question? The DICOM modalities I read from the test dataset are CT, MR, and MR. Why not CTA, MRA, and MRI? How can I obtain the correct modalities ?",
    "3296794": "MRA is a subset of MRI. In practice there is no straightforward way to distinguish any two given MRI sequences other than looking at the images and the header variables. Given the limited header information available in the test set, I suggest that you use the image data to determine what type of sequence you are dealing with."
  }
}