ISMRM-2024: NLCG-Net: A Model-Based Zero-Shot Learning Framework for Undersampled Quantitative MRI Reconstruction

11/4/2024 update: Our work has been accepted by ISMRM 2024 as poster! Link: https://arxiv.org/abs/2401.12004

Codes for T2 case have been optimised to be more compliant with the specification, making it easier to implement by specifying parameters in 'parser_ops.py'. Note that the work output is being developed into a journal paper, and the latest NLCG-Net turns out to be more powerful than when it was firstly purposed (08/11/2023).

1/2024 This version of code is being further organized so that it can be more reader-friendly. 't1' refers to T1 mapping, and 't2' refers to T2 mapping. Differences are mainly in 'data_consistency.py' because they have different imaging bases and raw data. Please feel free to comment if you meet any problem running codes