Construct deep neural networks based on direct sampling methods for solving electrical impedance tomography

Mathematics

• Electrical Impedance Tomography 100%
• Sampling Methods 76%
• Direct Method 66%
• Neural Networks 63%
• Ill-posedness 11%
• Efficacy 10%
• Computational Cost 9%
• Extremes 8%
• Cauchy 8%
• Inclusion 7%
• Inverse Problem 7%
• Numerical Experiment 6%
• Decompose 6%
• Demonstrate 5%

Engineering & Materials Science

• Deep neural networks 65%
• Tomography 65%
• Acoustic impedance 49%
• Sampling 47%
• Inverse problems 10%
• Convolutional neural networks 8%
• Decomposition 6%
• Neural networks 5%
• Costs 3%
• Experiments 3%