Segmentation

We have worked on a number of different general methods for image segmentation [1] [2] with a recent focus on segmentation methods based on convex relaxations which incorporate domain knowledge (for example in the form of encouraging directed orientation transitions [3] [4] [5]). We have also devised segmentation methods for the segmentations of fiber bundles from diffusion weighted images [6] as well as a method to extract connectivity information from diffusion tensor images [7] and have most recently explored a segmentation method which can impose constraints on the segmentation area [8].


References

  1. Zach C, Niethammer M, Frahm J-M
    2009.  Continuous Maximal Flows and Wulff Shapes: Application to MRFs. Proceedings of the Conference on Computer Vision and Pattern Recognition (CVPR). :1911–1918.
  2. Dambreville S, Niethammer M, Yezzi A, Tannenbaum A
    2007.  A variational framework combining level-sets and thresholding. :1–10.
  3. Zach C, Shan L, Frahm J-M, Niethammer M
    2009.  Globally Optimal Finsler Active Contours. DAGM. :552–561.
  4. Shan L, Zach C, Niethammer M
    2010.  Automatic three-label bone segmentation from knee MR images. International Symposium on Biomedical Imaging (ISBI).
  5. Shan L, Zach C, Styner M, Charles C, Niethammer M
    2010.  Automatic Bone Segmentation and Alignment from MR knee images. SPIE Medical Imaging.
  6. Niethammer M, Zach C, Melonakos J, Tannenbaum A
    2009.  Near-tubular fiber bundle segmentation for diffusion weighted imaging: segmentation through frame reorientation.. NeuroImage. 45(1 Suppl):S123-32.
  7. Niethammer M, Boucharin A, Zach C, Maltbie E, Shi Y, Styner M
    2010.  DTI Connectivity by Segmentation. MICCAI, International Workshop on Medical Imaging and Augmented Reality (MIAR).
  8. Niethammer M, Zach C
    2013.  Segmentation with area constraints.. Medical image analysis. 17(1):101-12.