Automatic CNN-based detection of cardiac MR motion artefacts using k-space data augmentation and curriculum learning

Keyphrases

• CNN-based 100%
• Motion Artifact 100%
• K-Space Data 100%
• Data Augmentation 100%
• Augmentation Learning 100%
• Cardiac MR 100%
• Curriculum Learning 100%
• Learning Approaches 50%
• Medical Image 50%
• UK Biobank 50%
• Good Quality 50%
• Cardiac Magnetic Resonance Image 50%
• Long-term Recurrent Convolutional Network 50%
• Temporal Convolutional Network 50%
• Population Data 25%
• Classification Performance 25%
• Quality Assessment 25%
• Image Quality 25%
• K-space 25%
• Network Architecture 25%
• Automatically Identify 25%
• Health Problems 25%
• Poor Quality 25%
• Cardiac Magnetic Resonance 25%
• Convolutional Neural Network Architecture 25%
• Deep Learning Architectures 25%
• Curriculum-based 25%
• UK Biobank Data 25%
• High Recall 25%
• Short Axis 25%
• Image Analysis pipeline 25%
• Quality Assessment Methods 25%
• Image Quality Assessment 25%
• Imbalanced Classification 25%
• Large Population Study 25%
• High-quality Imaging 25%
• Area under the ROC Curve 25%
• Manual Identification 25%
• Quality Assessment Techniques 25%
• Highly Imbalanced Classification 25%
• Cine Images 25%
• Cine Magnetic Resonance Imaging 25%

Computer Science

• Convolutional Neural Network 100%
• Data Augmentation 100%
• Learning Approach 66%
• Quality Assessment 66%
• Convolutional Network 66%
• Neural Network Architecture 33%
• Classification Problem 33%
• Image Quality 33%
• Image Analysis 33%
• Network Architecture 33%
• Image Quality Assessment 33%
• Classification Performance 33%
• Artifact Creation 33%
• Deep Learning Method 33%