{
  "id": 598636,
  "title": "Is it worth it converting DICOM to NIfTI?",
  "url": "/competitions/rsna-intracranial-aneurysm-detection/discussion/598636",
  "author_name": "Andres Felipe Romero Grajales",
  "post_date": "2025-08-12T05:24:42.859000",
  "votes": 0,
  "comment_count": 3,
  "views": 0,
  "content": "<p>I am currently in the preprocessing stage and considering whether to work directly with DICOM files using Pydicom or a similar library, or whether I should convert all the data to the “simpler” and more data-analysis-friendly NIfTI format, as discussed in other threads. I don’t have powerful computational resources, and an attempt to convert the series from DICOM to NIfTI caused my computer to crash. I wanted to ask the community, both organizers and fellow competitors, for their take on this issue. Is it worth converting all the multi-GB data into NIfTI for the competition, or is working directly with DICOM and Pydicom the better approach?</p>",
  "messages": [
    {
      "id": 3268621,
      "postDate": "2025-08-13T08:20:39.170Z",
      "content": "<p>I converted the dcms to .npy format.</p>",
      "rawMarkdown": "I converted the dcms to .npy format.",
      "votes": 1
    },
    {
      "id": 3268959,
      "postDate": "2025-08-13T16:56:03.060Z",
      "content": "<p>My take is to convert to a different format per my post on the topic. This conversion should not be highly resource intensive.</p>",
      "rawMarkdown": "My take is to convert to a different format per my post on the topic. This conversion should not be highly resource intensive."
    },
    {
      "id": 3268134,
      "postDate": "2025-08-12T13:04:20.303Z",
      "content": "<p>Convert to numpy I would say.</p>",
      "rawMarkdown": "Convert to numpy I would say."
    },
    {
      "id": 3267891,
      "postDate": "2025-08-12T05:24:42.860Z",
      "content": "<p>I am currently in the preprocessing stage and considering whether to work directly with DICOM files using Pydicom or a similar library, or whether I should convert all the data to the “simpler” and more data-analysis-friendly NIfTI format, as discussed in other threads. I don’t have powerful computational resources, and an attempt to convert the series from DICOM to NIfTI caused my computer to crash. I wanted to ask the community, both organizers and fellow competitors, for their take on this issue. Is it worth converting all the multi-GB data into NIfTI for the competition, or is working directly with DICOM and Pydicom the better approach?</p>",
      "rawMarkdown": "I am currently in the preprocessing stage and considering whether to work directly with DICOM files using Pydicom or a similar library, or whether I should convert all the data to the “simpler” and more data-analysis-friendly NIfTI format, as discussed in other threads. I don’t have powerful computational resources, and an attempt to convert the series from DICOM to NIfTI caused my computer to crash. I wanted to ask the community, both organizers and fellow competitors, for their take on this issue. Is it worth converting all the multi-GB data into NIfTI for the competition, or is working directly with DICOM and Pydicom the better approach?"
    }
  ],
  "comments": [
    {
      "id": 3268621,
      "author_name": "Seeing Times",
      "author_url": "",
      "post_date": "2025-08-13T08:20:39.170000",
      "content": "<p>I converted the dcms to .npy format.</p>",
      "votes": 1,
      "replies": []
    },
    {
      "id": 3268959,
      "author_name": "Evan Calabrese",
      "author_url": "",
      "post_date": "2025-08-13T16:56:03.060000",
      "content": "<p>My take is to convert to a different format per my post on the topic. This conversion should not be highly resource intensive.</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 3268134,
      "author_name": "jucious",
      "author_url": "",
      "post_date": "2025-08-12T13:04:20.303000",
      "content": "<p>Convert to numpy I would say.</p>",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "3268621": "I converted the dcms to .npy format.",
    "3268959": "My take is to convert to a different format per my post on the topic. This conversion should not be highly resource intensive.",
    "3268134": "Convert to numpy I would say.",
    "3267891": "I am currently in the preprocessing stage and considering whether to work directly with DICOM files using Pydicom or a similar library, or whether I should convert all the data to the “simpler” and more data-analysis-friendly NIfTI format, as discussed in other threads. I don’t have powerful computational resources, and an attempt to convert the series from DICOM to NIfTI caused my computer to crash. I wanted to ask the community, both organizers and fellow competitors, for their take on this issue. Is it worth converting all the multi-GB data into NIfTI for the competition, or is working directly with DICOM and Pydicom the better approach?"
  }
}