Explainable cardiac pathology classification on cine MRI with motion characterization by semi-supervised learning of apparent flow.
Qiao ZhengHervé DelingetteNicholas AyachePublished in: CoRR (2018)
Keyphrases
- semi supervised learning
- supervised learning
- improve the classification accuracy
- semi supervised classification
- wall motion
- cardiac imaging
- unsupervised learning
- magnetic resonance imaging
- semi supervised
- left ventricle
- labeled data
- unlabeled data
- co training
- machine learning
- transductive support vector machine
- labeled and unlabeled data
- semi supervised learning algorithms
- mr imaging
- text classification
- class labels
- mr images
- tagged mri
- label propagation
- manifold regularization
- classification accuracy
- unlabeled samples
- pattern recognition
- motion tracking
- mri images
- machine learning algorithms
- magnetic resonance images
- training data
- cardiac motion
- graph based semi supervised learning
- fully automatic
- active learning
- feature extraction
- feature space
- classification models
- magnetic resonance
- decision trees
- left ventricular
- beating heart
- data sets
- semi supervised learning methods
- short axis
- feature selection
- dimensionality reduction
- coronary artery
- label information
- object recognition
- patient specific
- metric learning
- pairwise
- learning tasks
- transfer learning
- high dimensional
- support vector machine