• I think I want to regroup orthogonal images from different DICON series

    From st2000@21:1/5 to All on Sat Jan 30 18:02:35 2021
    Hi,

    I have an MRI CD and am inspecting it using Envesalius in hopes of creating 3D printable objects. I am looking at 3 series. Each series has 1 normal resolution set of slices and 2 orthogonal low-resolution set of slices. None of these 3 series lend
    them selves to be rendered in 3D because of the 2 orthogonal low-resolution set of slices. But I believe if I could extract the normal resolution set of slices from each of the 3 series and create a new series I would finally be able to render 3D
    objects in Envesalius.

    But I don't know how to extract a specific set of slices from a series let alone how to create a new series.

    I am not even sure this approach will work.

    Any help would be appreciated.

    -thanks

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From David Gobbi@21:1/5 to All on Tue Feb 2 19:55:10 2021
    What you are thinking of is possible, but is difficult to do, and produces poor results. You can't simply combine series in different orientations into a single series and feed it into 3D software, since the software expects all slices in the series to
    have the same orientation. It is possible to resample each series to the same orientation and smaller slice spacing, and then combine them by averaging them or through more sophisticated techniques, but even then result will be far from the quality of
    the kinds of 3D MRI scans that are acquired for the purpose of 3D modelling.

    An MR scan protocol is designed according to the intended use of the scans, so your mileage may vary for any use other than the intended use!

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From st2000@21:1/5 to All on Wed Feb 3 12:26:33 2021
    Thank you for your response David. I have a suspicion that all 3 series are in fact the same data. I didn't want to include that as my description (to me) was already complex. Here's my hunch, I think in an effort to alleviate the processing load on
    the end-user's (in this case the doctor's) computer, 3 renderings of the MRI are created. But only 1 of the perspective in each rendering is of normal resolution. The other 2 are of low resolution and likely only give the doctor an idea of where he or
    she is in 3 dimensional space. Now the doctor can flip between series if they want to view an orthogonal normal resolution image. This is great for them as they can likely "move" through many slices of the normal resolution image very fast since the
    computer only needs to work hard on 1 of the 3 orthogonal images. But for people like me who need all 3 orthogonal images to be of normal resolution at the same time ... well, it's maddening to know you have all the data but can not put it to use.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From David Gobbi@21:1/5 to All on Wed Feb 3 13:06:31 2021
    The low-res scans might be reformats of the high-res scan, as you described, but they could just as easily be locator scans that the MR technician (not the physician) did first and used to decide exactly where to place the high-res scan slices.

    have all the data but can not put it to use.

    You have all the data that was collected, but it's a moot point since the amount of data collected was probably only 20% of what would be needed to build your 3D model (i.e. as much as 5 times as much scan time would have been needed). I can feel your
    pain, though.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)