{
  "id": 18053,
  "title": "Understanding the time metadata",
  "url": "/competitions/second-annual-data-science-bowl/discussion/18053",
  "author_name": "afterhours",
  "post_date": "2015-12-21T03:59:02.110000",
  "votes": 0,
  "comment_count": 0,
  "views": 576,
  "content": "<p>The Dicom files have quite a lot of metadata in them. I was hoping to use this metadata in some way. In particular, I wanted to understand the time information. </p>\n\n<p>One piece of metadata on there is AcquisitionTime, which I am assuming is the time at which the frame was taken.  I'm actually chiefly interested in the ordering, but I just want to make sure I have a correct understanding of the values in the fields.</p>\n\n<p>The Dicom time format seems to be &quot;hhmmss.microseconds&quot;. </p>\n\n<p>I loaded up all the saxs for study 1, and pulled the AcquisitionTime out of each sax.  Here is the time-between-collection I got:</p>\n\n<p>[datetime.timedelta(0, 27, 712500),\n datetime.timedelta(0, 33, 542500),\n datetime.timedelta(0, 25, 182500),\n datetime.timedelta(0, 35, 197500),\n datetime.timedelta(0, 25, 840000),\n datetime.timedelta(0, 30, 212500),\n datetime.timedelta(0, 48, 732500),\n datetime.timedelta(0, 25, 990000),\n datetime.timedelta(0, 26, 817500),\n datetime.timedelta(0, 26, 900000),\n datetime.timedelta(0, 36, 855000),\n datetime.timedelta(0, 24, 595000),\n datetime.timedelta(0, 250, 462500),\n datetime.timedelta(0, 837, 532500),\n datetime.timedelta(0, 124, 757500)]</p>\n\n<p>As you can see, the time delay is 25-50 seconds, but then there are three at the end which are much further separated in time.</p>\n\n<p>Is anyone else looking at the time information for stitching data? Can anyone shed any light on how to sequence the information cleanly in time?</p>",
  "messages": [
    {
      "id": 102264,
      "postDate": "2015-12-21T03:59:02.110Z",
      "content": "<p>The Dicom files have quite a lot of metadata in them. I was hoping to use this metadata in some way. In particular, I wanted to understand the time information. </p>\n\n<p>One piece of metadata on there is AcquisitionTime, which I am assuming is the time at which the frame was taken.  I'm actually chiefly interested in the ordering, but I just want to make sure I have a correct understanding of the values in the fields.</p>\n\n<p>The Dicom time format seems to be &quot;hhmmss.microseconds&quot;. </p>\n\n<p>I loaded up all the saxs for study 1, and pulled the AcquisitionTime out of each sax.  Here is the time-between-collection I got:</p>\n\n<p>[datetime.timedelta(0, 27, 712500),\n datetime.timedelta(0, 33, 542500),\n datetime.timedelta(0, 25, 182500),\n datetime.timedelta(0, 35, 197500),\n datetime.timedelta(0, 25, 840000),\n datetime.timedelta(0, 30, 212500),\n datetime.timedelta(0, 48, 732500),\n datetime.timedelta(0, 25, 990000),\n datetime.timedelta(0, 26, 817500),\n datetime.timedelta(0, 26, 900000),\n datetime.timedelta(0, 36, 855000),\n datetime.timedelta(0, 24, 595000),\n datetime.timedelta(0, 250, 462500),\n datetime.timedelta(0, 837, 532500),\n datetime.timedelta(0, 124, 757500)]</p>\n\n<p>As you can see, the time delay is 25-50 seconds, but then there are three at the end which are much further separated in time.</p>\n\n<p>Is anyone else looking at the time information for stitching data? Can anyone shed any light on how to sequence the information cleanly in time?</p>",
      "rawMarkdown": "The Dicom files have quite a lot of metadata in them. I was hoping to use this metadata in some way. In particular, I wanted to understand the time information. \r\n\r\nOne piece of metadata on there is AcquisitionTime, which I am assuming is the time at which the frame was taken.  I'm actually chiefly interested in the ordering, but I just want to make sure I have a correct understanding of the values in the fields.\r\n\r\nThe Dicom time format seems to be \"hhmmss.microseconds\". \r\n\r\nI loaded up all the saxs for study 1, and pulled the AcquisitionTime out of each sax.  Here is the time-between-collection I got:\r\n\r\n[datetime.timedelta(0, 27, 712500),\r\n datetime.timedelta(0, 33, 542500),\r\n datetime.timedelta(0, 25, 182500),\r\n datetime.timedelta(0, 35, 197500),\r\n datetime.timedelta(0, 25, 840000),\r\n datetime.timedelta(0, 30, 212500),\r\n datetime.timedelta(0, 48, 732500),\r\n datetime.timedelta(0, 25, 990000),\r\n datetime.timedelta(0, 26, 817500),\r\n datetime.timedelta(0, 26, 900000),\r\n datetime.timedelta(0, 36, 855000),\r\n datetime.timedelta(0, 24, 595000),\r\n datetime.timedelta(0, 250, 462500),\r\n datetime.timedelta(0, 837, 532500),\r\n datetime.timedelta(0, 124, 757500)]\r\n\r\nAs you can see, the time delay is 25-50 seconds, but then there are three at the end which are much further separated in time.\r\n\r\nIs anyone else looking at the time information for stitching data? Can anyone shed any light on how to sequence the information cleanly in time?"
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "102264": "The Dicom files have quite a lot of metadata in them. I was hoping to use this metadata in some way. In particular, I wanted to understand the time information. \r\n\r\nOne piece of metadata on there is AcquisitionTime, which I am assuming is the time at which the frame was taken.  I'm actually chiefly interested in the ordering, but I just want to make sure I have a correct understanding of the values in the fields.\r\n\r\nThe Dicom time format seems to be \"hhmmss.microseconds\". \r\n\r\nI loaded up all the saxs for study 1, and pulled the AcquisitionTime out of each sax.  Here is the time-between-collection I got:\r\n\r\n[datetime.timedelta(0, 27, 712500),\r\n datetime.timedelta(0, 33, 542500),\r\n datetime.timedelta(0, 25, 182500),\r\n datetime.timedelta(0, 35, 197500),\r\n datetime.timedelta(0, 25, 840000),\r\n datetime.timedelta(0, 30, 212500),\r\n datetime.timedelta(0, 48, 732500),\r\n datetime.timedelta(0, 25, 990000),\r\n datetime.timedelta(0, 26, 817500),\r\n datetime.timedelta(0, 26, 900000),\r\n datetime.timedelta(0, 36, 855000),\r\n datetime.timedelta(0, 24, 595000),\r\n datetime.timedelta(0, 250, 462500),\r\n datetime.timedelta(0, 837, 532500),\r\n datetime.timedelta(0, 124, 757500)]\r\n\r\nAs you can see, the time delay is 25-50 seconds, but then there are three at the end which are much further separated in time.\r\n\r\nIs anyone else looking at the time information for stitching data? Can anyone shed any light on how to sequence the information cleanly in time?"
  }
}