Statistical Shape Modeling Using MDL Incorporating Shape, Appearance, and Expert Knowledge
We propose a highly automated approach to the point correspondence problem for anatomical shapes in medical images. Manual landmarking is performed on a small subset of the shapes in the study, and a machine learning approach is used to elucidate the characteristic shape and appearance features at each landmark. A classifier trained using these features defines a cost function that drives key landmarks to anatomically meaningful locations after MDL-based correspondence establishment. Results are shown for artificial examples as well as real data.
- 1.Davies, R., et al.: A minimum description length approach to statistical shape modeling. IEEE TMI 21(5), 525–537 (2002)Google Scholar
- 2.Styner, M., et al.: Evaluation of 3D correspondence methods for model building. In: Taylor, C.J., Noble, J.A. (eds.) IPMI 2003. LNCS, vol. 2732, pp. 63–75. Springer, Heidelberg (2003)Google Scholar
- 3.Thodberg, H., et al.: Adding curvature to minimum description length shape models. BMVC 2, 251–260 (2003)Google Scholar
- 4.Heimann, T., et al.: Implementing the automatic generation of 3D statistical shape models with ITK. In: Open Science Workshop at MICCAI, Copenhagen (2006)Google Scholar
- 6.Manay, S., et al.: Integral invariants for shape matching. IEEE PAMI 28(10), 1602–1618 (2006)Google Scholar