{
  "id": 239992,
  "title": "Path to image",
  "url": "/competitions/siim-covid19-detection/discussion/239992",
  "author_name": "",
  "post_date": "2021-05-18T08:52:51.118304600Z",
  "votes": 1,
  "comment_count": 5,
  "views": 0,
  "content": "<p>Hi,</p>\n<p>in the description we can read: \"Note that all images are stored in paths with the form <strong>study/series/image</strong>. The study ID here relates directly to the study-level predictions, and the image ID is the ID used for image-level predictions.\"</p>\n<p><strong>study</strong> is the StudyInstanceUID of the train_image_level.csv.<br>\n<strong>image</strong> is included into the id of train_image_level.csv.</p>\n<p>But where can i extract the <strong>series</strong>???</p>\n<p>Best regards</p>",
  "messages": [
    {
      "id": "1312795",
      "postDate": "05/18/2021 08:52:51",
      "content": "<p>Hi,</p>\n<p>in the description we can read: \"Note that all images are stored in paths with the form <strong>study/series/image</strong>. The study ID here relates directly to the study-level predictions, and the image ID is the ID used for image-level predictions.\"</p>\n<p><strong>study</strong> is the StudyInstanceUID of the train_image_level.csv.<br>\n<strong>image</strong> is included into the id of train_image_level.csv.</p>\n<p>But where can i extract the <strong>series</strong>???</p>\n<p>Best regards</p>",
      "rawMarkdown": "Hi,\n\nin the description we can read: \"Note that all images are stored in paths with the form **study/series/image**. The study ID here relates directly to the study-level predictions, and the image ID is the ID used for image-level predictions.\"\n\n**study** is the StudyInstanceUID of the train_image_level.csv.\n**image** is included into the id of train_image_level.csv.\n\nBut where can i extract the **series**???\n\nBest regards",
      "votes": null
    },
    {
      "id": "1313020",
      "postDate": "05/18/2021 11:23:27",
      "content": "<p>Maybe do a listdir() in the study directory for sub directories, then look in the sub directory for the file.</p>",
      "rawMarkdown": "Maybe do a listdir() in the study directory for sub directories, then look in the sub directory for the file.",
      "votes": null
    },
    {
      "id": "1313034",
      "postDate": "05/18/2021 11:27:50",
      "content": "<p>Yes, i can see the folder and subfolder. But i need to know the series id for automation.</p>",
      "rawMarkdown": "Yes, i can see the folder and subfolder. But i need to know the series id for automation.",
      "votes": null
    },
    {
      "id": "1313081",
      "postDate": "05/18/2021 11:55:06",
      "content": "<p>Right, that's what I mean. Loop through the study_uids and check the study directories for sub directories, which are the series directories. Then store them in a dataframe. </p>",
      "rawMarkdown": "Right, that's what I mean. Loop through the study_uids and check the study directories for sub directories, which are the series directories. Then store them in a dataframe.",
      "votes": null
    },
    {
      "id": "1313094",
      "postDate": "05/18/2021 12:04:04",
      "content": "<p>Ok, this could be a solution. I thought there is a meaning of the series id. </p>",
      "rawMarkdown": "Ok, this could be a solution. I thought there is a meaning of the series id.",
      "votes": null
    },
    {
      "id": "1313120",
      "postDate": "05/18/2021 12:16:21",
      "content": "<p>The filenames aren't real DICOM UIDs. They look like hashes to me.</p>\n<p>They are stored in the DICOM files the same way. You could loop through all the directories and extract the data from the files.</p>\n<p>0020,000D = StudyInstanceUID = study folder<br>\n0020,000E = SeriesInstanceUID = series folder<br>\n0008,0018 = SOPInstanceUID = filename (without .dcm extension)</p>",
      "rawMarkdown": "The filenames aren't real DICOM UIDs. They look like hashes to me.\n\nThey are stored in the DICOM files the same way. You could loop through all the directories and extract the data from the files.\n\n0020,000D = StudyInstanceUID = study folder\n0020,000E = SeriesInstanceUID = series folder\n0008,0018 = SOPInstanceUID = filename (without .dcm extension)",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1313020,
      "author_name": "davidbroberts",
      "author_url": "",
      "post_date": "05/18/2021 11:23:27",
      "content": "<p>Maybe do a listdir() in the study directory for sub directories, then look in the sub directory for the file.</p>",
      "votes": null,
      "replies": [
        {
          "id": 1313034,
          "author_name": "drcapa",
          "author_url": "",
          "post_date": "05/18/2021 11:27:50",
          "content": "<p>Yes, i can see the folder and subfolder. But i need to know the series id for automation.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1313081,
          "author_name": "davidbroberts",
          "author_url": "",
          "post_date": "05/18/2021 11:55:06",
          "content": "<p>Right, that's what I mean. Loop through the study_uids and check the study directories for sub directories, which are the series directories. Then store them in a dataframe. </p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1313094,
          "author_name": "drcapa",
          "author_url": "",
          "post_date": "05/18/2021 12:04:04",
          "content": "<p>Ok, this could be a solution. I thought there is a meaning of the series id. </p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1313120,
          "author_name": "davidbroberts",
          "author_url": "",
          "post_date": "05/18/2021 12:16:21",
          "content": "<p>The filenames aren't real DICOM UIDs. They look like hashes to me.</p>\n<p>They are stored in the DICOM files the same way. You could loop through all the directories and extract the data from the files.</p>\n<p>0020,000D = StudyInstanceUID = study folder<br>\n0020,000E = SeriesInstanceUID = series folder<br>\n0008,0018 = SOPInstanceUID = filename (without .dcm extension)</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1312795": "Hi,\n\nin the description we can read: \"Note that all images are stored in paths with the form **study/series/image**. The study ID here relates directly to the study-level predictions, and the image ID is the ID used for image-level predictions.\"\n\n**study** is the StudyInstanceUID of the train_image_level.csv.\n**image** is included into the id of train_image_level.csv.\n\nBut where can i extract the **series**???\n\nBest regards",
    "1313020": "Maybe do a listdir() in the study directory for sub directories, then look in the sub directory for the file.",
    "1313034": "Yes, i can see the folder and subfolder. But i need to know the series id for automation.",
    "1313081": "Right, that's what I mean. Loop through the study_uids and check the study directories for sub directories, which are the series directories. Then store them in a dataframe.",
    "1313094": "Ok, this could be a solution. I thought there is a meaning of the series id.",
    "1313120": "The filenames aren't real DICOM UIDs. They look like hashes to me.\n\nThey are stored in the DICOM files the same way. You could loop through all the directories and extract the data from the files.\n\n0020,000D = StudyInstanceUID = study folder\n0020,000E = SeriesInstanceUID = series folder\n0008,0018 = SOPInstanceUID = filename (without .dcm extension)"
  },
  "source": "meta"
}