Imitation learning for improved 3D PET/MR attenuation correction
Medical Image Analysis(2021)
摘要
•CT-based losses neglect the PET data when doing MR to CT synthesis.•MR to CT synthesis networks should account for PET reconstruction error.•Multi-hypothesis uncertainty provides the best approximation of the true PET.•Imitation learning results in a better optima and more generalisable model.
更多查看译文
关键词
MR to CT synthesis,Deep learning,Imitation learning,Convolutional neural network
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要