Chapter 6 Osprey command line tutorial

All Osprey functions descibed in the GUI section can alternatively be called directly using a series of commands in the Matlab terminal. The function RunOspreyJob.m is a one-stop-shop wrapper for all of these commands, running the full analysis without interruption:

MRSCont = RunOspreyJob('Path/To/Job/File');

It may be preferable to view your analysis at each step by calling each command in turn. After setting up the job file, either as a Matlab .m file, .json, or .csv, the Osprey job is initialized as follows:

MRSCont = OspreyJob('Path/To/Job/File');

MRSCont will now contain the options and file locations for your data, as specified in the job file. Next, the data can be loaded:

MRSCont = OspreyLoad(MRSCont);

The substructure, MRSCont.raw, now contains the raw MRS data, as extracted from their native format. We may now preprocess these data:

MRSCont = OspreyProcess(MRSCont);

The MRSCont.processed will now contain the processed data, as well as some quality control measures in MRSCont.QM for your inspection. If all is well, we may now fit the data:

MRSCont = OspreyFit(MRSCont);

Modeling results are added to the MRSCont.fit, including results tables. If we have structural data, we can coregister our voxels, and perform tissue segmentation:

MRSCont = OspreyCoreg(MRSCont);
MRSCont = OspreySeg(MRSCont);

MRSCont now has sub-structures relating to voxel masks and volumes, MRSCont.coreg, and tissue segmentation results in MRSCont.seg.

Whether we had structural images or not, we may now quantify our spectra to produce some numerical results:

MRSCont = OspreyQuantify(MRSCont);

Finally, we may run the OspreyOverview function:

MRSCont = OspreyOverview(MRSCont);

This final function adds group measures, if we have a stat.csv file, and performs some statistics necesary for some of the plotting functions and GUI windows.

At any point during a command-line analysis, it is possible to visualize the data using the plot functions found in Osprey/plot, or even by switching to the GUI mid-analysis using:

OspreyGUI(MRSCont);