Weakly-Supervised Deep Learning for Left Ventricle Fibrosis Segmentation in Cardiac MRI Using Image-Level Labels.
Roel C. KleinFlorence E. van LieshoutMaarten Z. KolkKylian van GeijtenbeekRomy VosSamuel Ruiperez-CampilloRuibin FengBrototo DebPrasanth GanesanReinoud KnopsIvana IsgumSanjiv NarayanErik BekkersBob de VosFleur V. TjongPublished in: CinC (2022)
Keyphrases
- cardiac mri
- deep learning
- weakly supervised
- left ventricle
- image analysis
- superpixels
- physiological parameters
- segmented images
- segmentation method
- wall thickness
- segmentation algorithm
- fully automatic
- input image
- short axis
- image segmentation
- image features
- multiscale
- image data
- left ventricular
- topic models
- single image
- object class
- magnetic resonance images
- semi supervised
- test images
- machine learning
- image regions
- unsupervised learning
- object segmentation
- region of interest
- color histogram
- relation extraction
- high resolution
- viewpoint
- ground truth
- named entities