{
  "id": 18516,
  "title": "Systole and diastole volume calculation from series of 2D segmented image?",
  "url": "/competitions/second-annual-data-science-bowl/discussion/18516",
  "author_name": "",
  "post_date": "2016-01-22T12:00:57.267Z",
  "votes": null,
  "comment_count": 1,
  "views": 898,
  "content": "<p>Can anyone please explain me, how to calculate the volume of blood from segmented image? Say for example, if I have series of 2D segmented image (approx. 30 in the present case), then how can I calculate the systole and diastole volume?</p>\n\n<p>PS:\nI am a MATLAB user.</p>",
  "messages": [
    {
      "id": "105361",
      "postDate": "01/22/2016 12:00:57",
      "content": "<p>Can anyone please explain me, how to calculate the volume of blood from segmented image? Say for example, if I have series of 2D segmented image (approx. 30 in the present case), then how can I calculate the systole and diastole volume?</p>\n\n<p>PS:\nI am a MATLAB user.</p>",
      "rawMarkdown": "Can anyone please explain me, how to calculate the volume of blood from segmented image? Say for example, if I have series of 2D segmented image (approx. 30 in the present case), then how can I calculate the systole and diastole volume?\r\n\r\nPS:\r\nI am a MATLAB user.",
      "votes": null
    },
    {
      "id": "105404",
      "postDate": "01/22/2016 21:34:26",
      "content": "<p>If you manage to segment the individual images then the general strategy would be something like:</p>\n\n<ol>\n<li>Group the images by time. Typically there are 30 times for each location. </li>\n<li>For each time, integrate the areas at the different location slices to obtain a volume. You need to take into account the distances between the segmented slices to make the integral come out correctly. Also take into account the scale of the images.</li>\n<li>Choose the smallest and largest volumes for systole and diastole respectively.</li>\n<li>Convert the predicted volumes into CDFs</li>\n</ol>\n\n<p>I believe that the official tutorials actually do this after they segment the slices, so I recommend looking at those.</p>",
      "rawMarkdown": "If you manage to segment the individual images then the general strategy would be something like:\r\n\r\n1. Group the images by time. Typically there are 30 times for each location. \r\n2. For each time, integrate the areas at the different location slices to obtain a volume. You need to take into account the distances between the segmented slices to make the integral come out correctly. Also take into account the scale of the images.\r\n3. Choose the smallest and largest volumes for systole and diastole respectively.\r\n4.  Convert the predicted volumes into CDFs\r\n\r\nI believe that the official tutorials actually do this after they segment the slices, so I recommend looking at those.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 105404,
      "author_name": "bitsofbits",
      "author_url": "",
      "post_date": "01/22/2016 21:34:26",
      "content": "<p>If you manage to segment the individual images then the general strategy would be something like:</p>\n\n<ol>\n<li>Group the images by time. Typically there are 30 times for each location. </li>\n<li>For each time, integrate the areas at the different location slices to obtain a volume. You need to take into account the distances between the segmented slices to make the integral come out correctly. Also take into account the scale of the images.</li>\n<li>Choose the smallest and largest volumes for systole and diastole respectively.</li>\n<li>Convert the predicted volumes into CDFs</li>\n</ol>\n\n<p>I believe that the official tutorials actually do this after they segment the slices, so I recommend looking at those.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "105361": "Can anyone please explain me, how to calculate the volume of blood from segmented image? Say for example, if I have series of 2D segmented image (approx. 30 in the present case), then how can I calculate the systole and diastole volume?\r\n\r\nPS:\r\nI am a MATLAB user.",
    "105404": "If you manage to segment the individual images then the general strategy would be something like:\r\n\r\n1. Group the images by time. Typically there are 30 times for each location. \r\n2. For each time, integrate the areas at the different location slices to obtain a volume. You need to take into account the distances between the segmented slices to make the integral come out correctly. Also take into account the scale of the images.\r\n3. Choose the smallest and largest volumes for systole and diastole respectively.\r\n4.  Convert the predicted volumes into CDFs\r\n\r\nI believe that the official tutorials actually do this after they segment the slices, so I recommend looking at those."
  },
  "source": "meta"
}