BrainVoyager QX 2.1 Release Notes

New Features

GUI PluginsBrainVoyager QX 2.1 introduces an important extension to the scripting and plugin system allowing to create graphical user interface (GUI) components. The new extensions have been defined on the basis of a new scripting layer (Qt Script) that replaces the scripting module (QSA) of previous versions. While also a version of BrainVoyager QX 2.1 will be made available with the old scripting module, we strongly suggest to adapt existing scripts to the new scripting environment since only the new environment allows to use the implemented GUI functionality for scripts and plugins. The adaptation of scripts to the new system is very easy since the old and new scripting language are both based on (slightly different versions of) JavaScript. The examples in the "BVQXExtensions/Scripts" folder and the documentation "Scripting in BrainVoyager QX 2.1+" help to quickly make this transition. Scripts play an important role for GUI plugins since they allow to link actions on GUI elements (e.g. clicking a button) with appropriate calls into the plugin C++ code. Check the blog entry GUI Scripts and GUI Plugins as well as the "Writing GUI Plugins" documentation in the User's Guide to learn how to create GUI plugins.

 

Enhancements

MVPA - Extracting Trial ResponsesThe Multi-Voxel Pattern Analysis (MVPA) dialog fits a separate GLM for each trial of a time series in order to obtain a trial response estimate as a t or beta value. In some cases, it might be more appropriate to extract values directly without the involvement of a GLM. In this release, the "Extract Trial Data" dialog has been added allowing to extract raw data values in a window with respect to the time point of trial onset. If the window consists of one time point, a single value is extracted; if the window comprises 2 or more values, the mean of the values is extracted. If in addition a baseline interval is also specified (recommended), the extracted trial response value can be related to the trial baseline as a percent signal value or by subtracting the baseline from the raw trial response value. The new dialog can be invoked by clicking the "Options" button in the "Trial Estimation" tab of the "Multi-Voxel Pattern Analysis" dialog.
EMEG Suite - fMRI / EEG CouplingThe EMEG module now allows to explore the signal coupling of simultaneously recorded EEG and fMRI time-series allowing to predict fMRI signals directly from spatio-temporally selective EEG responses. Specifically, it is possible to extract regional power time-courses for fMRI modeling and prediction with appropriate options for the Source Analysis. For details, consult topic "EEG Source Analysis with EEG-fMRI Coupling" in the User's Guide.
ACPC, TAL Transformation of NR-VMPsWhile native-resolution VMPs have become the standard file format for almost any operation and the "anatomical-resolution" VMPs mostly serve visualization purposes now. in previous versions, it was, however, not possible to perform spatial transformations on native-resolution volume maps, but only on AR-VMPs. In order to obtain a NR-VMP in ACPC or TAL space, the time course (or diffusion-weighted) data had to be transformed first in that space followed by calculations leading to native resolution volume maps (recommended approach). Some users prefer, however, to calculate maps only in native space and would like to transform the resulting maps only into ACPC or Talairach space for visualization purposes and random effects analysis (e.g. of FA maps). The spatial transformation dialog ("Transform .VMP" button in "Spatial Transf" tab of "3D Volume Tools" dialog) as well as the Talairach and Un-Talairach dialogs ("ACPC -> TAL", "TAL -> ACPC" buttons in "Talairach" tab) have been extended allowing now to perform these transformations either on native-resolution (default) or anatomical-resolution VMPs.
Protocols with Parametric WeightsA new protocol version has been introduced (version 3) that allows to specify a weight for each defined interval. To add weights to a protocol, you need to turn on the "Enable parametric weights for intervals" flag in the "Stimulation Protocol" dialog. The specified weights are used in the "Single Study General Linear Model" dialog to scale intervals accordingly when defining predictors. When defining a predictor with non-constant trial weights, two predictors will be created: one predictor with weights "1.0" (identified as "Main" in the predictor name), and a second one having the amplitude modulated as indicated by the weights (identified as "parametric" in the predictor name). A new option field "Creation of parametric predictor" in the tab of the "Masking/Options" tab of the "Single Study GLM Options" dialog allows to decide whether the weights should be used as specified ("Use weights as defined" option), or whether the mean of the weights should be subtracted from each weight ("Subtract mean of weights" option, default) or whether the weights should be standardized ("Standardize weights" option). The latter two options are recommended since they ensure that the two defined predictors are orthogonal with respect to each other. The contribution of the parameric predictor can be tested in the "Overlay GLM Contrasts" dialog as usual by "plus selection"; it is recommended to perform a conjunction contrast of the two predictors (main and parametric). The implemented enhancement (changes in protocol and design matrix creation) makes it easier to specify parametric designs. In order to keep compatibility with previous protocol versions, a saved protocol is only stored as a new version 3 file if parametric weights are indeed used. In case parametric weights are not specified, protocols are still saved as version 2 PRT files allowing to use non-parametric protocols created in version 2.1 also in earlier versions of BrainVoyager.
Specification of Free Slice Scanning OrderWhile the standard slice order choices in the "FMR Preprocessing" dialog (ascending, descending, eventually combined with interleaved) are normally valid for most scans, there are cases that may have a special slice scanning order that is not captured by the default choices. For these cases, it is now possible to specify a free slice order. Furthermore, the specified slice order can be saved for later use. When running the slice scanning correction preprocessing step, the program now also prints detailed information about the performed temporal reordering of slices in the Log pane, also for the standard slice order choices.
Functional Connectivity AnalysisIt has been possible in previous versions to calculate correlation maps using a reference function extracted from a voxel or region-of-interest. Such correlation maps are very useful to show how the respective voxel or ROI is functionally connected to other voxels. In previous versions, this required to first save an extracted time course to disk, swithcing to the correlation (or GLM) dialog, loading the reference function and then executing the calculation. In this version, this function is now directly available by the "Run Correlation" command from the context menu of any (FMR-STC, VMR-VTC, SRF-MTC based) "ROI Signal Time Course" dialog.
ANCOVA DialogThe assignment of conditions was not updated in previous versions if within-factor levels has been deleted in the "Level list" of the "Design" tab of the "ANCOVA" dialog. This could lead to unexpected results when the user did not update the assignment in the "Assign Conditions" tab. In this release, the "auto-assign" functionality (available in the "Assign Conditions" tab) is now run automatically in the background when the user deletes factor levels in the "Design" tab. Other minor improvements have been made to increase overall stability of the ANCOVA module (e.g. the names of betas in the "ContrastPage" could be messed up in some cases).
Unified Color DialogIn previous versions, the definition of mesh colors (background color, curvature color) was available in two different dialogs. In this version, the "Background and Curvature Color" dialog has been added that centralizes all options to change mesh colors. This includes setting the background color as well as the two modes (2 colors mode, 2 x 2 colors gradient mode) to specify curvature colors.
Scripting CommandsThe scripting commands for creation of VTC and VDW data sets in native, ACPC and TAL space now support the new (5th) parameter "DataType" allowing to specify whether the resulting VTC/VDW files should contain float data (DataType = 2) or short integer data (DataType = 1). Note that existing scripts using one of the commands "CreateVTCInVMRSpace", "CreateVTCInACPCSpace", "CreateVTCInTALSpace" or the respective "CreateVDW" versions need to be updated. The "DataType" parameter must appear before the "Resolution" parameter. In order to support a more powerful creation of design matrices, the new properties "SDMContainsConstantPredictor" and "FirstConfoundPredictorOfSDM" have been added. The new properties "VMRVoxelResolutionX", "VMRVoxelResolutionY" and "VMRVoxelResolutionZ" can be used to read (and set) the resolution of VMR documents.
Cross in VMR DocumentsThe display and handling of the cross used to navigate in VMR documents has been improved in this release. The square in the center of the cross (VMR view) is now shown properly centered and its size may now be specified using the new "Rect. size" spin box in the "3D Coords" tab of the "3D Volume Tools" dialog. The square can also be hidden by setting the size value to "0".
Modeless VMR PropertiesSimilar as the "FMR Properties" and "DMR Properties" dialogs, the "VMR Properties" dialog has been made "modeless", i.e. it does no longer block inspection of documents when it is open.
Removing Non-Existing FilesA new "Clean" button has been added to the "Recent Files" pane allowing to remove those files from the recent files lists that are no longer availalbe on disk.

 

Bug Fixes

SDM Check for RFX-GLMsWhen running RFX-GLMs and a predictor (pooled across all runs of a subject) had only zero (or any constant) entries would lead to a RFX GLM result containing "not a number" (NAN) values for those predictors since no beta value can be estimated in that case. In order to avoid this issue and to inform the user about this problem, a consistency check of all subject-specific design matrices is now made when clicking the "GO" button in the "General Linear Model: Multi-Study, Multi-Subject" dialog. If (pooled) zero predictors are detected for a subject, an appropriate message is presented and the calculation is aborted.
Saving FMR/DMR FilesWhen a FMR or DMR project had been saved in previous versions, only the FMR/DMR text file itself had been saved but not the associated data files (STC/DWI, AMR, PRT). This issue has been fixed in this release. This also ensures that all associated files of a project are now stored in case that a FMR or DMR project is saved in a new folder.
Saving ResidualsWhen running a single-study GLM on a VTC data set, residuals were not saved when the "Save Residuals" option had been turned on in the "Serial Correlations Options" dialog. This problem has been fixed.
Multi-Run FMR GLMWhen running a multi-study GLM analysis on the basis of FMRs created in float format (multiple runs of the same session with the same slice positioning), the resulting maps were not calculated correctly. This problem has been fixed.
VMP Threshold SettingWhen the lower threshold of a VMP was set to value "0.0", the negative color (usually blue) disappered and only positive values were shown. This problem has been fixed in this release. Furthermore, values with a value of exactly 0.0 are no longer shown since they indicate missing values or indicate that a voxel is masked out.
Create VTC/VDW Memory LeakWhen creating VTCs or VDWs, part of the allocated memory was not free'd when the procedure was finished. This could lead to lack of memory in case that several VTCs had been createed in succession, e.g. when executing scripts to create multiple VTC/VDW files. This issue has been fixed.
ANCOVA - 1B1WWhen running a two-factorial ANCOVA with one between and one within factor, a scatterplot shows how the data of the different groups relates to the added covariate. In the previous version, the groups could not be distinguished since all data points received the same color. This issue has been fixed.
Transformation with Sinc InterpolationWhen running transformations with sinc interpolation that did not include rotations (e.g. ACPC -> TAL transformation) "garbage" could be obtained on some platforms (64-bit Mac). This problem has been fixed.
Lag-SMP from VMPWhen creating a surface map (SMP) from a volume map (VMP) containing lag values (e.g. from a cross-correlation analysis), the SMP did not contain lag values when this was done after loading the VMP to a VMR. It worked only when the VMP was calculated just before creating the SMP. This issue has been fixed.
Display of DMR-VMR TRF ParamsWhen performing a DMR-VMR coregistration, inspection of position info using the "Pos Info" button in the "Initial Alignment" tab of the "DMR-VMR Coregistration" dialog displayed wrong data. This issue has been fixed.
Restrict VTC Option and Mask CreationWhen mask files were created from cortex representations and the "Restrict within VTC boundaries" option was not checked in the "Create Cortex-Based VTC Mask" dialog, the program could crash. This issue has been fixed.
Cancel Mask CreationIf a VTC mask was planned to be created from a VMP but canceled when prompted to specify the MSK file name, the "busy" cursor was not reset. This problem has been fixed.
Axes and Lighting in Surface WindowWhen the display of axes was changed in the mesh display window (surface module), e.g. by pressing the "A" key repeatedly or by changing settings in the "3D Axes Options" dialog, the lighting could change on some platforms resulting in a "dark" mesh display. This issue has been fixed.
Switching Back from Other ProgramsWhen working in sub-dialogs of the "3D Volume Tools" dialog, especially when using the "Find AC Point" or "Find AC-PC Plane" dialogs, switching to other applications and then back to BrainVoyager could result in a crash on some platforms or disappearance of the sub-dialogs. This issue has been fixed.
Edit Condition Name in ProtocolWhen one would click the "Edit Name" button when no condition had been selected in the "Stimulation Protocol" dialog, the program would crash. This issue has been fixed.
Update of Cluster Threshold ChangeWhen one changed the cluster threshold value in the "Surface Maps" dialog, the display did not reflect the change always even when clicking the "Update SRF" button. This has been fixed now. In the updated program, ciclicking "Update SRF" is not strictly necessary - the display is now updated also when pressing "return" after editing the cluster threshold value.