Semi-supervised segmentation of echocardiography videos via noise-resilient spatiotemporal semantic calibration and fusion

Keyphrases

• Echocardiography 100%
• Semi-supervised Segmentation 100%
• Temporal Semantics 100%
• Semantic Fusion 100%
• Noise-resilient 100%
• Semantic Calibration 100%
• Feature Map 75%
• Speckle Noise 50%
• Irregular Motion 50%
• Endocardium Segmentation 50%
• Calibration Method 25%
• State-of-the-art Techniques 25%
• Clinical Practice 25%
• Adverse Effects 25%
• Fast Speed 25%
• Challenging Tasks 25%
• Video Data 25%
• Spatiotemporal 25%
• Current Frame 25%
• Left Ventricle 25%
• Limited Training Data 25%
• Unlabeled Data 25%
• Segmentation Performance 25%
• Temporal Perspective 25%
• Segmentation Accuracy 25%
• Dice Coefficient 25%
• Left Ventricular Function 25%
• Mean-teacher 25%
• Anisotropic Mobility 25%
• Annotation Cost 25%

Computer Science

• Feature Map 100%
• Annotation 33%
• Unlabeled Data 33%
• Segmentation Performance 33%
• Neighboring Frame 33%
• segmentation accuracy 33%
• Training Data 33%
• Temporal Information 33%
• Consecutive Frame 33%