Fine-tuning deep learning model parameters for improved super-resolution of dynamic MRI with prior-knowledge
作者：
Highlights：
• Alleviates the spatio-temporal trade-off in dynamic MRI using super-resolution
• Model was first trained on a static dataset with different contrast and resolution.
• Static subject-specific planning scan was used to fine-tuned the trained model.
• Achieved a structural similarity of 0.957 while reconstructing 6.25% of the k-space
摘要
•Alleviates the spatio-temporal trade-off in dynamic MRI using super-resolution•Model was first trained on a static dataset with different contrast and resolution.•Static subject-specific planning scan was used to fine-tuned the trained model.•Achieved a structural similarity of 0.957 while reconstructing 6.25% of the k-space
论文关键词：Super-resolution,Dynamic MRI,Prior knowledge,Fine-tuning,Patch-based super-resolution,Deep learning
论文评审过程：Received 23 April 2021, Revised 7 October 2021, Accepted 12 October 2021, Available online 15 October 2021, Version of Record 26 October 2021.
论文官网地址：https://doi.org/10.1016/j.artmed.2021.102196
