Multi-class prediction for improving intestine segmentation on non-fecal-tagged CT volume.
Hirohisa OdaYuichiro HayashiTakayuki KitasakaAitaro TakimotoAkinari HinokiHiroo UchidaKojiro SuzukiMasahiro OdaKensaku MoriPublished in: Medical Imaging: Computer-Aided Diagnosis (2022)
Keyphrases
- multi class
- ct volume
- ct images
- x ray
- ground truth
- medical images
- segmentation algorithm
- multi class classification
- prediction accuracy
- pairwise
- object detection
- multiple classes
- support vector machine
- feature selection
- intraoperative
- cost sensitive
- multi class classifier
- image segmentation
- medical imaging
- multi class svm
- binary classification
- energy function
- computed tomography
- multi class boosting
- multiscale
- level set
- segmentation method
- base classifiers
- shape prior
- machine learning
- protein classification
- binary classifiers
- statistical shape model
- ct data
- multi class problems
- data sets
- error correcting output codes
- image processing
- ct scans
- automatic segmentation
- training data
- prior information
- image analysis