Comparison of Neural Network Architectures for Physics-Driven Deep Learning MRI Reconstruction

Mathematics

• Network Architecture 100%
• Neural Networks 61%
• Learning 59%
• Data Consistency 57%
• Physics 54%
• Machine Learning 36%
• Feedforward 21%
• Iteration 21%
• Mask 20%
• Architecture 18%
• Optimization Algorithm 17%
• Iterative Procedure 17%
• Alternate 16%
• Encoding 16%
• Iterative Algorithm 15%
• Least Squares 14%
• Regularization 14%
• Strategy 11%
• Unit 11%
• Model 10%
• Performance 9%
• Operator 7%

Engineering & Materials Science

• Magnetic resonance imaging 69%
• Network architecture 61%
• Deep learning 58%
• Physics 56%
• Machine learning 48%
• Neural networks 39%
• Masks 31%

Business & Economics

• Deep Learning 90%
• Neural Networks 78%
• Physics 74%
• Machine Learning 43%
• Regularization 21%
• Least Squares 17%
• Data Base 14%
• Operator 14%
• Performance 6%