Fully convolutional multi-scale residual DenseNets for cardiac segmentation and automated cardiac diagnosis using ensemble of classifiers.
Mahendra KhenedAlex VargheseGanapathy KrishnamurthiPublished in: Medical Image Anal. (2019)
Keyphrases
- multiscale
- cardiac mri
- heart disease
- blood pool
- cardiac mr images
- clinically relevant
- patient specific
- cardiovascular disease
- ensemble learning
- left ventricle
- cardiac magnetic resonance
- segmentation algorithm
- training data
- tagged mri
- ensemble pruning
- ensemble classifier
- image segmentation
- training set
- multiple classifiers
- fully automatic
- majority voting
- multiple scales
- ensemble methods
- feature selection
- coronary artery
- neural network
- medical diagnosis
- edge detection
- image analysis
- support vector
- fully automated
- wall motion
- classifier ensemble
- image representation
- automated segmentation
- single photon emission computed tomography
- echocardiographic images
- computed tomography angiography
- spect images
- machine learning methods
- medical images
- learning algorithm
- normal subjects
- final classification
- coarse to fine
- training samples
- natural images
- scale space
- multi class
- decision trees