Segmentation of 3D Medical Objects by Deformable M-reps

 [Last update: 23-Oct-2000 by tracton@radonc.unc.edu]

Papers

The concepts and methods for this segmentation technique are described in the paper

Pizer SM, Fletcher PT, Fridman Y, Fritsch DS, Gash AG, Glotzer JM, Joshi S, Thall A, Tracton G, Yushkevich P, Chaney EL, (2000). Deformable M-Reps for 3D Medical Image Segmentation. Submitted for publication to MedIA. [Adobe Acrobat Reader PDF format]

Further geometric and proabilistic theory on m-reps can be found in

Pizer SM, Fritsch DS, Yushkevich P, Johnson V, Chaney EL (1999). Segmentation, Registration, and Measurement of Shape Variation via Image Object Shape. IEEE Transactions on Medical Imaging, 18 (10): 851-865.  [Postscript format, PDF format coming soon...]

Additional Results

These are results beyond those which can be found in the first paper, and most are under construction

Movies are presented in AVI and RealPlayer formats, easily viewed on Windows or Macintosh platforms:

  • AVI - for both Microsoft's Windows Media Player and Apple's Quicktime Player (standalone player or browser plugin; Windows/Mac). You need to wait for the entire movie to be copied to your local disk before viewing can begin. You might prefer this if you have a fast (10MBit/sec) network connection. Modern computers usually come with these programs pre-installed. Both are free.
  • RealPlayer is required to view a movie as it is copied. Although the advantages are that you can start viewing the movie much more quickly and you do not need to store the movie on your local disk (we enabled storing in case you need it), the downside is that the quality may be poorer than the AVI version in some cases. The server adapts the image quality to match your current connection speed, which can vary continuously. Get the free basic player, version 6 or above.

No audio is used, so you won't disturb the person sitting in the next airplane seat. :)

Kidney Results

  1. Cine of the model deformation (segmentation) sequence to extract a kidney from a 3D CT data set:

    The kidney model, a single figure modeling the kidney parenchyma plus the renal pelvis     AVI MOVIE , RealPlayer
  2. Cine views of image slices vs. segmented kidneys for results which segmented successfully fully automatically after hand placement of model.

    Stage of Progress

    Example 1

    Example 2

    Example 3

    Example 4

    Example 5

    During hand placement
    After hand placement
    After all m-rep stages (similarity transform, boundary displacement)
    After the boundary displacement, i.e.,
      at the end of the segmentation

  3. Cine of image slices vs. segmented kidneys for results which converged incompletely under automatic operation after hand placement of model

    Stage of Progress

    Example 1

    Example 2

    Example 3

    Example 4

    After hand placement
    After all m-rep stages (similarity transform, boundary displacement)

    After repeating automatic segmentation with
      the previous automatic result as the model
            

Cerebral Ventricle Results

TBD

Forming a model

TBD