Skip to content

Conversation

HJZollner
Copy link
Collaborator

Merging latest develop into the master branch

-Added MRSI mask for fitting
- New visualization for MRSI data for processed and fitted data
- data/list loading for arbitrary average numbers
- fixed MoCo bug related to the MRSi implementation where the phase between on and off is 180 degree
- added fitting of water data for MRSI
Fix MoCo and add water fit
- updated PE-edited MP basis sets (added EA basis function)
- delete Philips 40 ms unedited basis sets which was not stored in the right folder
- added Philips 80 and 140 ms unedited basis sets
- Removed H2O from the default metabolite list, so that it is not included in the metabolite fit anymore. It can still be included if the user specifies H2O in his metabolite list manually.
- The water basis function is loaded from the specific basis set if it is needed for the fit of the water scan. This happens regardless of the metabolite list the user specifies.
-adding bet for masking
-updating quantify MRSI to allow metabolite maps from different sub-spectra
-Apparently the index of the number kx and ky lines have changed in R 5.7.1
- Include a water ref normalization
- Added flags to control the thresholds and region for the MoCo script
- Added flags to control the subspectra alignment
- Added a zero line and more infos to the MRSIprocess plot
- Added the option to plot the uncorrected data to the processed view
- Added processing of the data that is not motion corrected (this still needs a flag to be deactivated later)
- Fixed segmentation
Added a new Osprey module OspreyMinReport.m that creates a markdown table including the minimum parameters for reporting of MRS studies (defined in MRSinMRS). This includes the following changes to other functions:

- Extracting software versions during the loading process (io_loadspec_sdat, io_loadspec_twix, and io_loadspec_GE)
- Automatic generation of md report during OspreyOverview call
Initial_commit_MRSinMRS_markdown
- Bug fix related to vendor-specific naming conventions in the voxel geometry
- Minor layout  changes
- Updated the calculation of the fit quality parameter to match the original paper Barros & Slotboom. The quality is defined as a variance ratio between residual in the fitted region and noise
Two minor bug fixes:
1) Possible issues in the metabolite list when new metabolites are added to a default Osprey basis set

2) Avoid crashes when hidden files are stored in the DCM folder
-Added a decision tree, that creates a metabolite list for the preliminary reduced fit according to the input list of the user (e.g. if the user performs a phantom scan with fCho instead of GPC or PCh in the list) this metabolite is picked instead of the default (NAA, Cr, PCh, Glu, and Ins.
Fixes reported Bug in the correlation plot within the GUI where it was not possible to switch between external variables
HJZollner and others added 29 commits May 11, 2021 17:28
-Updated functionalities for multi slice MRSI data
This file doesn't belong here.
Add a different route to retrieve the FID from Siemens DICOM files.
Updating loader for 3 slice MRSI data
Updates for 3 slice MRSI
- aded HERMES/HERCULES to the MRSinMRS report function
-updated the Siemens HERCULES I basis set to have a correct MM09 in the diff1 spectrum
- added MM3co model to the conc model
- updated the basis set for HERCULES I Siemens
Some updates based on the Jim's feedback:

- 40 ms unedited Siemens basis set
-  VE11 leftshift bug fixed
- OspreyOverview can be performed on data that is not modeled to generate scaled spectra and mean plots in the OverviewTab
Changed the control file to defaults settings and added the LCModel freeware key.
One addtional change to match teh LCM default
Fix GE-DICOM-based voxel coregistration.
Saving some space by creating a generic function checking that all modules required to run a particular module have been executed.

Also some small clean-ups here and there.

Renaming folder 'utilites' to 'utilities'.
- Changed the default model for modeling metabolite-nulled data. The reference shift is estimated from the MM data now, while FWHM estimates are still taken from the metabolite data. This model performs better than the older version (fund empirically)
- Fixes bug related to not being able to create overview spectra from the metabolite nulled data
- Updated example jobFiles to follow BIDS structure for automated generation.

- Examples for a direct path to the data are also included

- The version check is only giving a soft warning and is not stopping the GUI from working

- Renamed the Gauss fixed model to reflect the paper naming

- Updates to the jobFiles to reflect the most recent changes (e.g. no HERCULES1/2 and new metabolites as editing targets)

-added new Version number for release v.1.0.2
@HJZollner HJZollner merged commit f91c703 into master Jul 9, 2021
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants