MRIReco.jl: An MRI Reconstruction Framework written in Julia

• DOI: 10.48550/arxiv.2101.12624
• Publication Date: 2021-01-01
• Authors: Tobias Knopp; Mirco Grosser
• Keywords: Toolbox; Modular programming; Implementation; Interface (matter)

Purpose: The aim of this work is to develop a high-performance, flexible and easy-to-use MRI reconstruction framework using the scientific programming language Julia. Methods: Julia modern, general purpose with strong features in area signal / image processing numerical computing. It has high-level syntax but still generates efficient machine code that usually as fast comparable C/C++ applications. In addition itself, sophisticated package management system makes proper modularization functionality across different packages feasible. Our developed MRIReco.jl can therefore reuse existing from other concentrate on MRI-related parts. This includes common imaging operators support for raw data formats. Results: simple use high degree accessibility. While providing simple-to-use interface, many its components easily be extended customized. performance compared Berkeley Advanced Reconstruction Toolbox (BART) we show achieves speed popular library. Conclusion: Modern languages bridge gap between accessible implementations. leverages fact contributes promising environment future algorithmic development reconstruction.