Two-Stage Structure-Focused Contrastive Learning for Automatic Identification and Localization of Complex Pelvic Fractures

Keyphrases

• Accurate Diagnosis 33%
• Attention Mechanism 33%
• Attention Module 33%
• Automatic Diagnosis 33%
• Automatic Identification 100%
• Automatic Localization 100%
• Challenging Tasks 33%
• Classification Network 33%
• Clinical CT 66%
• Complex Pelvic Fracture 100%
• Complex Structure 33%
• Comprehensive Experiment 33%
• Computed Tomography 33%
• Contrastive Learning 100%
• CT Images 33%
• Feature Difference 33%
• Focus of Attention 33%
• Fracture Zone 100%
• Fragment Size 33%
• Geometric Features 33%
• Image Features 33%
• Irregular Morphology 33%
• Large Fragment 33%
• Learning Methods 33%
• Left-handed 33%
• Multiple Fragments 33%
• Multiple Sites 33%
• Pelvic Anatomy 33%
• Pelvic Bones 33%
• Pelvic Fracture 100%
• Publicly Available Dataset 33%
• Recognition Ability 33%
• Right-sided 33%
• Severe Trauma 33%
• Siamese Deep Neural Network 33%
• Size Difference 33%
• Spatial Structural 33%
• Structural Attention 33%
• Structural Features 33%
• Supervised Contrastive Learning 33%
• Surgical Planning 33%
• Symmetric Feature 33%
• Symmetry Property 33%
• Two-stage Structure 100%
• Two-weight 33%

Biochemistry, Genetics and Molecular Biology

• Computer Assisted Diagnosis 100%
• X-Ray Computed Tomography 100%