BrainVoyager QX 2.8 Release Notes

New Features

Population Receptive Field (pRF) Estimation This version introduces the population receptive field (pRF) estimation method. Using appropriately described stimuli, the model-driven pRF method estimates the location and receptive field size of each voxel or vertex included in the analysis. Since eccentricity and polar angle maps can be derived from estimated pRF parameters, the pRF method is usually used as an alternative to conventional phase-encoded retinotopic mapping but it provides more information such as estimates of the population receptive field size. For more details how to use the method for volume (VMR-VTC) and surface (SRF-MTC) data, consult the "Population Receptive Field Estimation" chapter of the User's Guide.
Multi-Dimensional Similarity Scaling (MDSS)

Multi-dimensional similarity scaling (MDSS) is a statistical tool to visualize the (dis)similarities between pairs of objects in which the objects are represented as points in a low dimensional space and the dissimilarities as distances between these points. The visualization in the low (e.g. 2) dimensional space is performed under the constraint that the distances between points correspond as closely as possible to the dissimilarities between the objects in original (high-dimensional) feature space. With the new MDSS plugin it is now possible to calculate and create such visualizations where the objects correspond to spatially distributed brain activity patterns originating from several fMRI observations in a given anatomical space, e. g. multi-voxel patterns in volume space (volume maps) or multi-vertex patterns in cortical surface space (surface maps). The interactive visual tool provides many new possibilities, e.g. it helps to find outliers and it allows to explore whether maps from different subjects form a single cluster or whether they form separate clusters; in a cognitive experiment, the latter case could potentially indicate that subjects have used different cognitive strategies to perform the "same" task. For further details, consult the documentation provided by the plugin (select the help link in the short description for the MDSS plugin available under "Plugins - Description Of Plugins" menu).

Multiple Linear Regression

… For further details, consult the documentation provided by the plugin (select the help link in the short description for the MLR plugin available under "Plugins - Description Of Plugins" menu).

Adaptive Map Smoothing

… For further details, consult the documentation provided by the plugin (select the help link in the short description for the MLR plugin available under "Plugins - Description Of Plugins" menu).




Enhancements


          

Automatic Advanced Segmentation In previous versions of BrainVoyager, two segmentation pipelines were available. The standard automatic segmentation pipeline was used to obtain topologically correct cortex meshes for many purposes, including advanced 3D visualiation and cortex-based alignment. The advanced segmentation pipeline was used to perform cortical thickness analyses. The standard segmentation pipeline operated at 1mm resolution VMRs (in ACPC or Talairach space) while the advanced segmentation operated at sub-millimeter (usually 0.5mm) resolution. These two formerly separate analyses are now integrated, i.e. it is possible to run the automatic segmentation pipeline using the routines previously only available for advanced segmentation. Furthermore, the advanced segmentation dialog now contains a new tab allowing to create cortex segmentations in both sub-millimeter and standard (1mm) VMR resolution. This allows to optionally use the advanced segmentation dialog to fine-tune processing options if needed. Furthermore, the segmentation routines have been further improved (see also next two points), including anatomical preprocessing (e.g. ventricle removal) at any (sub-)millimeter resolution. For further details, consult the User's Guide...
Advanced Mesh Smoothing When segmented cortical hemispheres are reconstructed, a mesh smoothing step is usually applied in order to remove the "jags" stemming from the voxelated initial surface representations. The standard smoothing has, however, the disadvantage that mesh vertices are displaced inwards, i.e. the depths of gyri and sulci tend to "shrink" and the resulting smoothed mesh no longer represents the originally reconstructed volume segmentation accurrately. The newly introduced advanced mesh smoothing routine avoids this problem, i.e. smoothing is restricted to high-frequencies such as the jags resulting from voxel reconstructions. The new "reco" smoothing has been built into the improved standard and advanced segmentation pipelines and it is also available in the "Meshes - Advanced Mesh Smoothing" menu.
Resolution-Independent Talairach Processing In previous version Talairach coordinates could be defined only on 1 mm (256 dimensions) VMR data sets (with limited support for 0.5 mm VMRs). It is now possible to define Talairach coordinates and to run Talairach transformations on any high-resolution VMR data sets. It is now also possible to create VTC files in ACPC and Talairach space with transformation and Talairach landmark files defined on sub-millimeter VMRs. Furthermore, Talairach grids are visualized correctly at any resolution and Talairach coordinates are converted to system coordinates and vice versa at any voxel resolution. The new possibility is also built into the "Advanced Segmentation" dialog (see also above), which accepts data sets in ACPC and Talairach space with any (sub-millimeter) resolution. For further details, consult the topic "Creation of High-Resolution VTCs" in the "Tools for High-Resolution Data" chapter of the User's Guide.
Mesh Scripting The available scripting commands have been substantially extended in this release. The most substantial change is the addition of a "mesh" object that allows to call, among others, mesh morphing, CBA and statistical routines. A number of new example scripts are provided (file names start with the "Mesh" substring) demonstrating how to use the new scripting possibilities. For further details, consult the...
Plugin API The BVQX Plugin interface API now supports access to GLM data structures. Furthermore the API functions for data plotting have been extended, including the possibility to display provided axis labels. For further details, consult the updated BrainVoyager QX Developer's Guide. In order to use the new functions, the latest API include files need to be downloaded from the plugins web site.
Improved High-Res Grid Sampling Export The "High-Resolution Cortex Grid Sampling" dialog now supports more convenient export options for sampled surface map data. Futhermore, it is now possible to export grid-sampled time course (VTC) data in addition to volume map (VMP) data. For details, see the updated documentation in the User's Guide.
GPU-Based Sinc Interpolation The GPU-based sinc interpolation routine introduced in the previous version for anatomical and map transformations is now also available for converting FMR data to VTC data in native, ACPC [todo: and Talairach] space. If enabled in the "Preferences" dialog, the fast routine is automatically used when selecting sinc interpolation in the "Create VTC Options" dialog. Furthermore, spatial transformation of V16 data now also uses GPU-based sinc interpolation if enabled.
Threshold Map Delegation In the context of "Preference" maps it became possible to visualize the content of a map that is thresholded by another map. This possibility has now been extended from preference maps to any combination of map types, i.e. it enables visualization of information from a selected map that is thresholded by another map (containing e.g. t or correlation values). This threshold map delegation approach is, for example, used in the context of derived pRF retinotopic maps to visualize angle maps that are thresholded by a separate multiple correlation map. The feature can be turned on manually by adding to a map name the string "UseThreshMap: " followed by the name of the thresholding map. It is also possible to create a new version of a map that is filtered by the thresholding of another map; in a filtered map, all values not passing the threshold of the referenced thresholding map are set to 0.0. This filtering option can be applied using the new "Filter UseThreshMaps" button available in the "Combine VMPs" and "Combine SMPs" dialogs. For further details, consult the topic "PRF Estimation in Cortical Surface Space" of the User's Guide.
Angle Maps In the context of pRF estimation, a new "angle" map type ("a") has been introduced in order to appropriately represent, process (e.g. spatially smooth) and visualize (polar) angle maps. Furthermore, to facilitate the mapping of displayed map colors to the represented angle, a filled circle (disk) is now displayed in volume and surface windows in addition to the standard map bar. For further details, consult the topic "PRF Estimation in Cortical Surface Space" of the User's Guide.
Brain Peeler The Brain Peeler has been improved providing robust results. The version in the "Main " tab of the "Volume Tools" dialog now allows to run only the first steps of the tool that mainly involve a high-pass filter to reduce intensity inhomogeneities without actually removing tissue. Running the tool with the "Apply only high-pass filter" option is recommended when one wants to reconstruct the skin of a participant's head.
Image Reporter The Image Reporter tool stores relevant image information when executing various routines, including automatic Talairach transformation and pRF estimation. Besides the previously available possibility to save individual frames as images to disk, Image Reporter now allows exporting all available images as a standard movie file.




Bug Fixes



Reaaply Spatial Transformation The reapply spatial transformation operation (available in the "VMR Properties" dialog) did not work if started from a VMR in Talairach space, i.e. the resulting VMR contained only empty (0-value) voxels. Furthermore, additional reapplication operations could not be launched sinde the "Reapply" button was disabled after using it once. These issues have been fixed.
To-SAG Transformation When applying a "To-SAG" transformation, the VMR data is not transformed by a standard transformation-with-interpolation routine but the axes of the VMR are simply changed accordingly; the resulting fast 90 degree rotation transformation has the advantage that the voxel values are not changed. Also the dimensions are only "rotated" but not changed (e.g. to 256). In case of non-isovoxel data sets, the voxel dimension were, however, not reassigned to the new axes in previous versions. This issue has been fixed.
GLM Settings When running GLMs, some settings such as serial correlation was sometimes turned off when overlaying contrasts (especially beta maps). When re-running the GLM, it looked as if the results have changed but the difference was related to the different GLM settings. This issue has been fixed, i.e GLM settings are not changed by subsequent "Overlay" functions.
Philips REC/PAR on Linux When creating projects from Philips REC/PAR files on case-sensitive file systems (e.g. Linux and some Mac systems), the .PAR filecorresponding to a selected .REC file can sometimes not be located even if it exists. The reason is that BrainVoyager looked only for lower-case file extensions (e.g. ".par") but not for upper-case (e.g. ".PAR") extensions. This issue has been fixed.
DICOM on Mac OS X When creating projects from DICOM files on Mac OS X, the program would hang when looking for files in a folder that does not contain any files; this was due to an issue with the reporting dialog that could not be launched. This issue has been fixed.
Adding ArVMPs on Mac When adding anatomical-resolution VMPs on Mac after loading the first ArVMP from the native resolution dialog, the program produced a crash. This issue has been fixed.
Readability of texts in surface window When visualizing surface maps, text labels depicting threshold values and statistical information (e.g. t, p values) were not visible when using a white background color. This has been fixed by ensuring that the text color always differs from the chosen background color.
Quality of Snapshot Export The quality hint in the "Bitmap" tab of the "Preferences" is used when saving snapshots of windows and dialogs to disk in .jpeg and .tiff formats. In case that not the default setting was used, the value selected by the quality slider was not intepreted correctly. This issue has been fixed.