Fully Convolutional Multi-scale Residual DenseNets for Cardiac Segmentation and Automated Cardiac Diagnosis using Ensemble of Classifiers.
Mahendra KhenedAlex VargheseGanapathy KrishnamurthiPublished in: CoRR (2018)
Keyphrases
- multiscale
- heart disease
- cardiac mri
- cardiac mr images
- patient specific
- blood pool
- left ventricle
- image segmentation
- clinically relevant
- cardiac magnetic resonance
- ensemble learning
- fully automatic
- training data
- feature selection
- training set
- classifier ensemble
- multiple scales
- neural network
- edge detection
- spect images
- echocardiographic images
- ensemble pruning
- medical diagnosis
- image analysis
- wall motion
- segmentation algorithm
- single photon emission computed tomography
- level set
- ensemble classifier
- left ventricular
- automated segmentation
- coronary artery
- weak classifiers
- tagged mri
- naive bayes
- learning algorithm
- majority voting
- left atrium
- decision trees
- normal subjects
- short axis
- final classification
- cardiovascular disease
- accurate classifiers
- support vector
- support vector machine
- fully automated
- ensemble methods
- coarse to fine
- computer aided
- segmentation method
- class labels
- image representation