Accurate localisation of the prostate and its surrounding tissue is essential in the treatment of prostate cancer.
This paper presents a novel approach to fully automatically segment the prostate, including its seminal vesicles, within a few minutes of a Magnetic Resonance (MR) scan acquired without an endorectal coil. Such MR images are important in external beam radiation therapy, where using an endorectal coil is highly undesirable. The segmentation is obtained using a deformable model that is trained on-the-fly so that it is specific to the patients scan. This case specific deformable model consists of a patient specific initialised triangulated surface and image feature model that are trained during its initialisation. The image feature model is used to deform the initialised surface by template matching image features (via normalised cross-correlation) to the features of the scan. The resulting deformations are regularised over the surface via well established simple surface smoothing algorithms, which is then made anatomically valid via an optimised shape model. Mean and median Dices Similarity Coefficients (DSCs) of 0.85 and 0.87 were achieved when segmenting 3T MR clinical scans of fifty patients. The median DSC result was equal to the inter-rater DSC and had a mean absolute surface error of 1.85 mm. The approach is showed to perform well near the apex and seminal vesicles of the prostate.
Written by:
Chandra S, Dowling J, Shen K, Raniga P, Pluim J, Greer P, Salvado O, Fripp J. Are you the author?
Reference: IEEE Trans Med Imaging. 2012 Aug 2. Epub ahead of print.
doi: 10.1109/TMI.2012.2211377
PubMed Abstract
PMID: 22875243
UroToday.com Prostate Cancer Section
