Weakly non-rigid MR-TRUS prostate registration using fully convolutional and recurrent neural networks.
Qiulan ZengYabo FuJiwoong Jason JeongYang LeiTonghe WangHui MaoAshesh B. JaniPretesh PatelWalter J. CurranTian LiuXiaofeng YangPublished in: Medical Imaging: Image Processing (2020)
Keyphrases
- recurrent neural networks
- prostate biopsy
- prostate segmentation
- image guided
- transrectal ultrasound
- image guided radiotherapy
- magnetic resonance
- prostate cancer
- registration accuracy
- patient specific
- image registration
- mr images
- imaging artifacts
- medical imaging
- neural network
- target registration error
- feed forward
- reservoir computing
- echo state networks
- recurrent networks
- prostate brachytherapy
- surface registration
- mr imaging
- automated segmentation
- artificial neural networks
- intraoperative
- medical images
- radiation therapy
- accurate registration
- nonrigid registration
- deformable registration
- texture information
- deformable models
- nonlinear dynamic systems
- treatment planning
- thin plate spline
- minimally invasive
- robot assisted
- imaging modalities
- contrast enhanced
- registration process