2009 Prostate segmentation challenge MICCAI

Goal

To discuss the state-of-art segmentation of Prostate MRI in the context of MRI-guided prostate therapy, through comparison of the segmentation methods using sample data.

Organizer

• Nobuhiko Hata (co-lead organizer, BWH, Boston, MA)
• Gabor Fichtinger (co-lead organizer, Queens Univ, Kingston, ON)
• Sota Oguro (Data managing, BWH, Boston, MA)
• Haytham Elhawary (Slicer module for validation, BWH, Boston, MA)

Rule

1. Download 3D Slicer for viewing and manipulating the dataset we provide.
2. (Optional) Download the tutorial and its associated images regarding prostate image processing from Slicer's website.
3. Download the Prostate MRI and the manual segmentation (as training data) from the website of National Center for Image Guided Therapy (PI: Jolesz and Tempany).
4. Download the Prostate MRI of the challenge subject.
5. Submit segmentation results with the paper describing the method developed and applied. The segmentation result should be NRRD format and loadable to Slicer. The segmentation result should aligned to original MRI when loaded to Slicer.
6. segmentation result should maintain the same grayscale value assinged to the training dataset. Namely,
7. the organizer will score the segmentation result by comparing it with manual segmentation of the expert using the announced comparison scheme.
8. Repeat step above at the workshop.

Dates

1. Training data available: May, 2009

1. Submission of results: June 29, 2009

1. Submission of papers (with results): July 20, 2009

1. Workshop: September 2009

Download

1. 3D Slicer
2. Prostate MRI tutorial for 3D Slicer
3. | Training MRI data and segmentation(N=15)
4. Subject data (for paper submission)
5. Subject data (for contest on-site)

Evaluation scheme

• 95% Hausdorff distance (HD)

The HD between two point datasets formed by the edges of the segmented prostate sub-anatomy from the MRI and manual segmentation. We will measured the accuracy of alignment between these two segmentations, by extracting the edges from the subsections. The HD is the maximum distance of a set to the nearest point in the other set. More formal description of the HD can be found at:

Archip N, Clatz O, Whalen S, et al. Non-rigid alignment of pre-operative MRI, fMRI, and DT-MRI with intra-operative MRI for enhanced visualization and navigation in image-guided neurosurgery. Neuroimage 2007; 35:609-624.

• Dice Similarity Coefficient (DSC)

The DSC, which has been used in evaluation of segmentation, gives a measure of the volumetric overlap between the segmented livers from the registered images. The DSC indicates twice the number of voxels which are shared by or are common to both registered datasets. The denominator is the total number of voxels in both datasets. The DSC can range from zero to one, where zero is no alignment between registered livers and one is perfect alignment. An example of DSC used in prostate segmetation can be found at

A. Bharatha, M. Hirose, N. Hata, S. K. Warfield, M. Ferrant, K. H. Zou, E. Suarez-Santana, J. Ruiz-Alzola, A. D'Amico, R. A. Cormack, R. Kikinis, F. A. Jolesz, and C. M. Tempany, "Evaluation of three-dimensional finite element-based deformable registration of pre- and intraoperative prostate imaging," Med Phys, vol. 28(12), pp. 2551-60, 2001.