updates 2016-01 January-April

Owned by Paul Macey (admin)
May 03, 2016
5 min read
2016-04-29
  • Added further automation to FSL FIRST processing (running the designs).
2016-04-24
  • Tidied up some ASL file checking.
2016-04-22
  • Lots of physiology work.
2016-04-19
  • Modified "skip file check" option to be Never overwrite - ON/Off; set in Analysis GUI. Also have option to not apply T1 rotations to ASL, DTI, fMRI, T2 - see debug box in analysis GUI.
2016-04-14
  • ASL - added filter option to threshold and median filter CBF maps
  • DTI - tidied up some code
2016-04-12
  • Fixed bug where creating summary images led to stuff being written in command window - should have been written to scripts that can be re-run to create check-reg figures.
2016-04-06
  • Minor tweaks to make code smoother.
2016-04-05
  • EEG export - which click EEG table, if EEG export option in CSPM is on, will save plots to Excel.
2016-04-04
  • Fixed fMRI (possible ASL) bug that would appear sometimes - to do with old physiology settings being mixed with SPM12
  • Remove a couple of unused bottoms from Project GUI, hide one unused setting in Local Settings GUI (physiology)
  • Deleted a couple of obsolete files ("mrilink")
  • Fixed recently-introduced DTI bug
2016-04-01
  • Tweak T1 rotations for fMRI, ASL, DTI, T2 - odd issue where there is a delay in the filesystem noting when the rotations were applied to the file, which confuses the timing check for when parameters were applied. This led to repeating application of rotations and hence forcing re-preprocessing.
  • Tried to fix an possible bug where new files were not copied to stats folders - not sure if this is one-off or not. Did for ASL CBF, need to check for other types.
2016-03-31
  • PAR/REC - no longer rescales image intensities according to scaling factor in PAR file; this scaling factor was sometimes ~1000, resulting in intensity values in the 100,000 to 500,000 range, which causes problems for processing.
    • Will this have negative consequences?
  • Fixed new ASL bug.
  • Fixed bug when adding comment in Sets.
2016-03-29
  • Update: T2 and DTI skip file check should work now.
  • Fixed database bug recently introduced - would get stuck whenever called something from database (recursive loop).
  • T1 rotations applied to original fMRI/DTI/ASL/T2 files
  • Option to skip checking for newer files  (working with fMRI only at present)
  • Ability to create 3D images like this: 
2016-03-27
  • Option to extract all individual VOI from a label file with multiple VOI (Under VOI menu).
  • Fixed MRIcroS error when reading very large VOI file.
2016-03-26
  • Image info - added sum of non-zero voxels (for calculating sum of residuals).
2016-03-25
  • More tweaks on shape surface display, im_fill function.
2016-03-23
  • More tweaks on shape surface display.
2016-03-22
  • Refined FIRST shape display GUI - code smoother, shape color option, show 0/1/2 overlays.
  • Tidied FIRST setup to avoid directory remove warning.
2016-03-21
  • Fixed bug in set filter GUI which would occur after selecting matching subjects and clicking Excel option.
2016-03-20
  • Tided batch file for Linux creation - now in temp rather than LinuxShare folder.
  • Tried to reduced command window text for check-reg when creating summaries.
  • Fixed fMRI bug with T1-to-MNI rotations (applied too early).
2016-03-18
  • Bleeding edge! T1-to-MNI rotations are now applied to DTI, ASL, fMRI, and T2 scans. This is not 100% tested so use carefully. Also this will re-preprocess everything.
  • FIRST - if no designs are present, gives a tidier exit and opens a how-to page (used to give an error).
2016-03-16
  • Added couple of steps in FIRST preprocessing to try and avoid errors due to system being too fast
  • Fixed bug in fMRI VOI combine repeat challenges timetrend option which would give a too-small baseline for some longer challenges.
  • Added some tweaks in fMRI VOI display to try and avoid crashes with Excel (perhaps code is running too fast, so added delays)
  • The T1-to-MNI parameters are now saved in the database. Thus if the analysis folder gets deleted, CSPM will apply existing parameters (if they exist) when the same study is preprocessed. 
  • The T1-to-MNI rotations are applied to ASL data; this makes the ASL-to-native-T1 registration, and MNI normalization, much more likely to be successful
    • Still need to apply this to other modalities: fMRI, DTI, T2
  • Made more changes to MRIcroS so that now two overlays with different color maps can be added to a single surface
2016-03-10
  • AAL pre-defined VOI now works for timetrends/structural values (in Display results).
  • Manual T1-to-MIN parameters are now saved in database; this is the first step in apply rotations to DTI, T2, other modalities.
  • Created GUI to show FIRST hippocampus and other surface overlays. Also modified MIRcroS matlab files.
  • Working on T1R, without success so far.
2016-02-28
  • Fixed DICOM convert bug for ASL / DTI series (probably was recently introduced - caused DICOM rename/convert to crash).
2016-02-26
  • Tidied some filtering in Edit groups for sets GUI (especially for empty values)
  • Doing some work on creating nice hippocampus overlays with MRIcroS.
2016-02-18
  • FSL FIRST - working on display with MRIcroS; new "fill surface" option for hippocampus (+others) masks (CSPM im menu). 
2016-02-17
  • EEG plots - next attempt at exporting. 
2016-02-16
  • SPM12 update (patches from SPM); lots of SPM files affected. 
2016-02-14
  • Found CSPM Excel sheet issue, changed so will be notice in command window, but not crash or give dialog box warning (to do with Excel 31 character limit on worksheet name)
  • Visualization with orthviews (CSPM, Visualization, multiple views) - now works correctly with blobs (colormap or single colors); need to also update spm file spm12\spm_orthviews, not just CSPM files.
2016-02-11
  • Added check for error when attempting to name an Excel worksheet; will give message, but not crash. 
  • Fixed fMRI temporal smoothing bug.
2016-02-09
  • Fixed bug in adding study to project group (only came up on some computers)
  • Figure printing with orthviews (under CSPM, Visualization menu) now works correctly and without loss (SPM/MATLAB update had caused errors and poor quality images). 
    • Still need to update all other CSPM printing functions.
  • Need to install Ghostscript for printing to work. (http://ghostscript.com/download/gsdnld.html)
  • Manually, can use "export_fig' tool to create figures; see README in CSPM\export_fig\ subfolder.
2016-02-04
  • Conversion of Philips T1 relaxometry (may still need to get TR)
  • Minor adjustments to analysis - trying to figure out why does not have detrending method sometimes
  • FreeSurfer tweaks
  • Tidied RFx code.
2016-01-27
  • Fixed mrVista-induced error that occasionally occurred during preprocessing.
2016-01-26
  • Fixed "archive" bug in Sets GUI
  • Added parallel processing check in a Linux call
2016-01-24,25
  • Fixed bug for studies with more than one T1 series, where an error would occur if the first of two or more T1's was missing files.
  • More physiology
  • Added FreeSurfer check for smoothed surface files; if missing, run recon-all with -qcache option (no idea what this does, but followed suggestion from email  list); also check for "IsRunning" flag.
2016-01-22
  • More physiology
2016-01-19
  • Physiology stuff; going through RMANOVA steps
2016-01-19
  • Physiology stuff; working through the programming required (linking phys to MRI, displaying in project configuration, running phys analysis)
2016-01-17
  • Fixed minor bug where set name was not saved if it had some non-allowable characters in name (in Preprocess).
2016-01-16
  • Can add and edit multi-line text in subject GUI for comment and medical history
  • Tidying error messages on covariates GUI
  • More work on physiology (converting old data, configuration GUI)
2016-01-10
  • More background changes for physiology - table for linking phys and mri, add phys asks for mri to link (for fMRI phys types).
  • Describe table tweak
2016-01-07
  • GE DTI: worked with using header information, now should work even if that option is not selected (goes to backup).
2016-01-02
  • GE DTI: some files were not being read because the DICOM DTI information was in different fields. The SPM dicom dictionary is updated, and the CSPM conversion now reads the b-value and directions.