One of the main reasons why MRS is still not widely used for multi-site clinical studies is the enormous diversity in data processing methods. Virtually every lab has tailor-made 'in-house' code that they have sometimes cultivated for decades. Unfortunately, MRS metabolite concentration estimates are extremely sensitive to details of data processing, modeling, and quantification workflows.
This project aims to create a standardized library of basic and advanced data processing classes in the most widely used programming languages (Python, MATLAB, R, C++). The foundational building blocks of this library will be the basic MRS processing steps (zero-filling, line-broadening, concatenating and splitting transients, etc.). As the library grows, more complex methods will be incorporated.
We will build all classes around the new NIfTI-MRS data storage specification to ensure optimal interoperability and translatability of processing workflows.