Denoising diffusion-based MR to CT image translation enables whole spine vertebral segmentation in 2D and 3D without manual annotations.
Robert GrafJoachim SchmittSarah SchlaegerHendrik Kristian MöllerVasiliki Sideri-LampretsaAnjany SekuboyinaSandro Manuel KriegBenedikt WiestlerBjoern H. MenzeDaniel RueckertJan Stefan KirschkePublished in: CoRR (2023)
Keyphrases
- medical images
- ct images
- denoising
- magnetic resonance
- x ray images
- medical imaging
- computed tomography
- mr images
- mr imaging
- inter patient
- ct and mr images
- anatomical structures
- medical image analysis
- ct volume
- ct scans
- deformable models
- region of interest
- image registration
- computer tomography
- x ray
- image denoising
- total variation
- lymph nodes
- brain images
- accurate segmentation
- imaging modalities
- magnetic resonance imaging
- image processing
- anisotropic diffusion
- noisy images
- segmentation algorithm
- level set
- treatment planning
- brain tumors
- clinical applications
- shape prior
- region growing
- ct data
- automated segmentation
- patient specific
- radiation therapy
- image data
- image analysis
- three dimensional
- magnetic resonance images
- ground truth
- cone beam
- image segmentation