Selection of Optimal Hyper-Parameters for Estimation of Uncertainty in MRI-TRUS Registration of the Prostate.
Petter RisholmFirdaus JanoosJennifer PursleyAndriy FedorovClare M. TempanyRobert A. CormackWilliam M. Wells IIIPublished in: MICCAI (3) (2012)
Keyphrases
- prostate biopsy
- hyperparameters
- image guided
- closed form
- prostate segmentation
- transrectal ultrasound
- model selection
- mr imaging
- magnetic resonance
- prostate cancer
- cross validation
- prostate brachytherapy
- bayesian inference
- sample size
- em algorithm
- mr images
- bayesian framework
- image registration
- random sampling
- prior information
- intraoperative
- medical imaging
- deformable models
- incomplete data
- noise level
- magnetic resonance images
- missing values
- worst case
- automatic segmentation
- incremental learning
- medical image analysis
- patient specific
- registration accuracy
- maximum likelihood
- support vector