Output structure attributes

Back to top

  The fields listed on this page are applicable to releases 3.3.0+ and are not necessarily correct for other versions of Gannet.

All output from Gannet analyses is saved in a structure (e.g., MRS_struct, or whatever name the user has chosen) that contains many fields and subfields. Listed below are lists of these fields and their descriptions. Note that this is not an exhaustive list of attributes of output that can be generated by Gannet.

MRS_struct

Field	Description
loadtime	The time when GannetLoad was initiated
version	Structure containing release numbers of Gannet and the Gannet modules
ii	During analysis, the current file being loaded; after analysis, the number of files that were loaded
metabfile	Cell array containing the inputted metabolite data filenames
waterfile	Cell array containing the inputted water reference data filenames

MRS_struct.p

Field	Description
target	The metabolite(s) targeted by the editing pulses
ON_OFF_order	Order of the editing subexperiments
seqorig	Origin of the Philips MEGA-PRESS or GE HERMES sequences
LB	Exponential line-broadening applied (Hz)
water_ECC	Whether eddy-current correction was applied to the water data
metab_ECC	Whether eddy-current correction was applied to the metabolite data
water_removal	Whether the residual water signal in DIFF spectrum was removed using HSVD
alignment	Choice of shot-to-shot frequency-and-phase correction
use_prealign_ref	Whether RobustSpecReg used the averaged pre-aligned subspectra as references to align the averaged post-aligned subspectra
fit_resid_water	Whether the residual water signal in the OFF spectrum was fitted to calculate a water suppression factor
weighted_averaging	Whether weighted signal averaging was implemented
HERMES	Whether the data are HERMES data
HERCULES	Whether the data are HERCULES data
PRIAM	Whether the data are PRIAM data
phantom	Whether the data are phantom data
join	Whether the input files were joined
mat	Whether the output structure was saved as MAT-file
csv	Whether a CSV file of useful data was created
append	Whether the PDF outputs were appended into single files
hide	Whether the output figures were hidden
vendor	Scanner vendor
reference	Concentrations will be calculated relative to this; if a water reference is provided, concentrations will be calculated relative to both water and Cr, otherwise they will be calculated relative to Cr
numScans	Total number of scans in the batch
numFilesPerScan	When input files are joined, the number of files joined per scan
npoints	Number of data points of each input file
nrows	Number of data frames stored in each input file
nrows_water	Number of data frames stored in each input water reference file
Navg	Total number of averages collected at acquisition
Nwateravg	Total number of averages collected at acquisition
TR	Repetition time (ms)
TE	Echo time (ms)
TR_water	Repetition time (ms)
TE_water	Echo time (ms)
LarmorFreq	Larmor frequency (MHz)
sw	Spectral width (Hz)
voxdim	Array of voxel dimensions (mm)
voxoff	Array of voxel position offset (mm)
voxang	Array of voxel angulation (deg)
ZeroFillTo	Processed spectra are zero-filled to this many data points
zf	Zero-fill factor
dt	Dwell time (s)
SpecRes	Spectral resolution of the raw data (Hz/point)
SpecResNominal	Nominal spectral resolution of the processed data (Hz/point)
Tacq	Acquisition time (s)
weighted_averaging_method	The algorithm used for weighted signal averaging

MRS_struct.fids

Field	Description
data	Raw time-domain metabolite data (dimensions: npoints \(\times\) nrows) of the last loaded dataset
data_water	Raw time-domain water reference data (dimensions: npoints \(\times\) nrows_water) of the last loaded dataset
ON_OFF	Array of the editing order of the last loaded dataset (1 = ON; 0 = OFF)
data_align	Raw time-domain metabolite data of the last loaded dataset with frequency-and-phase correction applied

MRS_struct.spec

Field	Description
vox1 (or whatever was set in MRS_struct.p.vox in GannetPreInitialise.m)	This structure contains the complex frequency-domain spectral data for all input datasets
freq	Frequency axis (ppm), calculated from the last loaded dataset
F0freq	Cell array of the observed frequency of either residual water signal or the 3 ppm Cr signal for each average
F0freq2	Cell array of the observed frequency of the 3 ppm Cr signal for each average; this is used in RobustSpectralRegistration.m as the starting values for frequency during optimization
AllFramesFT	Spectra of the last loaded dataset without frequency-and-phase correction applied
AllFramesFTrealign	Spectra of the last loaded dataset with frequency-and-phase correction applied

MRS_struct.out

Field	Description
AvgDeltaF0	Vector containing a metric of the amount of frequency offset observed during the acquisition, calculated as the mean difference between the observed frequency of the residual water signal (or the 3 ppm Cr signal) in the pre-frequency-corrected subspectra and the nominal water frequency at 4.68 ppm (or the nominal 3 ppm Cr signal if HERMES or GSH editing)
SpecReg	Structure containing output from spectral registration
reject	Cell array of binary integers indicating whether an individual transient should be rejected during signal averaging
signal_averaging	Structure containing a cell array of the weighting factors applied if weighted signal averaging is used
vox1 (or whatever was set in MRS_struct.p.vox in GannetPreInitialise.m)	This structure contains the output from the signal modeling, quality analysis, and quantification of all the signals of interest
QA	Structure containing metrics calculated to quantify the performance of the tissue segmentation

MRS_struct.mask

Field	Description
outfile	Cell array of the output filenames of the voxel masks
img	Cell array containing three-dimensional arrays representing the three orthogonal slices of the T1-weighted structural images that are displayed in the figure outputs
T1image	Cell array of the input T1-weighted structural image filenames
img_montage	Cell array containing three-dimensional arrays representing a montage of axial slices of the T1-weighted structural images overlaid by the tissue-segmented voxel masks that are displayed in the GannetQuantify.m figure output
mask_overlap	Structure giving details of where a voxel mask overlap is saved (this is created if the normalize flag is set to 1 in GannetPreInitialise.m)

---

Built with R Markdown in RStudio

Copyright © 2020–2026, Mark Mikkelsen