DMC-Net: Lightweight Dynamic Multi-scale and Multi-resolution convolution network for pancreas segmentation in CT images

Keyphrases

• Computed Tomography Images 100%
• Multi-resolution 100%
• Pancreas Segmentation 100%
• Multi-scale Resolution 100%
• Convolution Network 100%
• Pancreas 50%
• Multi-scale Dynamics 50%
• Multi-scale Convolution 50%
• Convolution Module 50%
• Multi-scale Convolution Module 37%
• Dice Similarity Coefficient 25%
• Dynamic Mechanism 25%
• Convolutional Neural Network 25%
• Global Context Information 25%
• Kernel Size 25%
• Lightweight Design 25%
• Organ Size 12%
• Interdependency 12%
• State-of-the-art Techniques 12%
• Organ Shape 12%
• Abdominal Computed Tomography 12%
• Medical Image Segmentation 12%
• Low Computational Complexity 12%
• Design-led 12%
• U-Net 12%
• Computational Complexity 12%
• Convolutional Neural Network Architecture 12%
• Inter-subject Variation 12%
• Segmentation Performance 12%
• Extracting Features 12%
• Segmentation Network 12%
• Convolutional Layer 12%
• Segmentation Accuracy 12%
• Global Information 12%
• Drop-in Replacement 12%
• Local Receptive Field 12%
• Convolutional Filter 12%
• Long-range Contextual Information 12%
• Large Kernel 12%

Computer Science

• Convolution 100%
• Computer Assisted Tomography 100%
• Convolutional Neural Network 16%
• Contextual Information 16%
• Multi-Scale Convolution Module 16%
• Computational Complexity 11%
• Image Segmentation 5%
• U-Net 5%
• Neural Network Architecture 5%
• Segmentation Performance 5%
• Network Segmentation 5%
• Convolution Layer 5%
• Global Information 5%
• segmentation accuracy 5%
• Receptive Field 5%
• Convolutional Filter 5%