SPINEPS - Automatic Whole Spine Segmentation of T2-weighted MR images using a Two-Phase Approach to Multi-class Semantic and Instance Segmentation.
Hendrik Kristian MöllerRobert GrafJoachim SchmittBenjamin KeinertMatan AtadAnjany SekuboyinaFelix StreckenbachHanna SchönFlorian KoflerThomas KrönckeStefanie BetteStefan WillichThomas KeilThoralf NiendorfTobias PischonBeate EndemannBjoern H. MenzeDaniel RueckertJan S. KirschkePublished in: CoRR (2024)
Keyphrases
- multi class
- mr images
- accurate segmentation
- medical images
- mr imaging
- partial volume
- brain tumors
- brain mr images
- mr brain
- magnetic resonance
- fully automatic
- magnetic resonance images
- brain structures
- level set
- bias field
- segmentation algorithm
- lesion segmentation
- cardiac magnetic resonance
- tissue segmentation
- intensity inhomogeneity
- image segmentation
- intensity distribution
- dice similarity coefficient
- mri data
- multiscale
- mr brain images
- manual segmentation
- inhomogeneity correction
- pairwise
- energy function
- shape prior
- automatic segmentation
- model based segmentation
- multi class classification
- multiple classes
- object segmentation
- contrast enhanced
- support vector machine
- knn
- nonrigid registration
- feature selection