manual for batchprocessingwizard for 4d data in brainvoyager€¦ · 4-dimensional data processing...

Manual for batchprocessingwizard

for 4D data in BrainVoyager

Copyright 2018 c© Brain Innovation B.V.

Contents

1 Introduction 11.1 Notes per wizard version . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.1.1 Notes on version 1.9 for BrainVoyager 20.6 . . . . . . . . . . . . . . . . . . . . . . . . . 11.1.2 Notes on version 1.4 for BrainVoyager QX 2.4 . . . . . . . . . . . . . . . . . . . . . . . 11.1.3 Notes on version 1.3 for BrainVoyager QX 2.4 . . . . . . . . . . . . . . . . . . . . . . . 11.1.4 Notes on version 1.2 for BrainVoyager QX 2.4 . . . . . . . . . . . . . . . . . . . . . . . 11.1.5 Notes on version 1.0 for BrainVoyager QX 2.2 . . . . . . . . . . . . . . . . . . . . . . . 11.1.6 Notes on version 0.8 for BrainVoyager QX 2.1 . . . . . . . . . . . . . . . . . . . . . . . 11.1.7 Notes on version 0.5 for BrainVoyager QX 2.1 . . . . . . . . . . . . . . . . . . . . . . . 11.1.8 Notes on version 0.6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.1.9 Notes on version 0.5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.1.10 Notes on version 0.4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.1.11 Notes on version 0.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31.1.12 Notes on version 0.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

1.2 Installation of the wizard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41.2.1 Location of the plugin and script files . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

1.3 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

2 Launching the wizard and global options 62.1 Description of the functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62.2 Global options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82.3 Processing order . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

2.3.1 Order of processing of data sets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102.3.2 Order of processing steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

2.4 Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102.4.1 BrainVoyager Log tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

3 Creating FMR or DMR projects 113.1 Specifying parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

3.1.1 Creating FMR projects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113.1.2 Creating DMR projects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113.1.3 Storage order of the slices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123.1.4 Set project parameters manually . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

4 Mean intensity adjustment 144.1 Specifying parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144.2 Background: mean intensity adjustment in BrainVoyager . . . . . . . . . . . . . . . . . . . . . 14

4.2.1 From User’s Guide of BrainVoyager QX 1.8 . . . . . . . . . . . . . . . . . . . . . . . . . 14

2

5 Slice scan time correction 155.1 Specifying parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

5.1.1 Slice order . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155.1.2 Interpolation method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

5.2 Background: slice scan time correction in BrainVoyager . . . . . . . . . . . . . . . . . . . . . . 175.2.1 From User’s Guide of BrainVoyager QX 1.9 . . . . . . . . . . . . . . . . . . . . . . . . . 175.2.2 From User’s Guide for BrainVoyager 2000 . . . . . . . . . . . . . . . . . . . . . . . . . 19

6 Performing motion correction and intra-session alignment 236.1 Specifying parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236.2 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

7 Temporal filtering 257.1 Specifying parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257.2 Background: temporal filtering in BrainVoyager . . . . . . . . . . . . . . . . . . . . . . . . . . 26

7.2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267.2.2 High pass filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267.2.3 Gaussian temporal smoothing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

8 Spatial filtering 288.1 Specifying parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288.2 Background: spatial smoothing in BrainVoyager . . . . . . . . . . . . . . . . . . . . . . . . . . 29

8.2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

9 Coregistration 309.1 Specifying parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309.2 Background: coregistration (and spatial normalisation) in BrainVoyager . . . . . . . . . . . . 30

10 Creating VTC/VDW files 3210.1 Specifying parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

10.1.1 File type for normalisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3210.1.2 Parameters for VTC/VDW creation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3210.1.3 Target space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3310.1.4 Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3310.1.5 Interpolation type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3310.1.6 Bounding box determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

11 Preprocessing VTC files 3411.1 Specifying parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

11.1.1 VTC spatial smooting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3411.1.2 VTC temporal high-pass filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3411.1.3 VTC temporal low-pass filtering (smoothing) . . . . . . . . . . . . . . . . . . . . . . . . 34

12 Creating MTC files 3612.1 Specifying parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3612.2 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

13 File selection 3813.0.1 Selecting files manually . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4013.0.2 Selection of files via a file list . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

3

Abstract

Outline This document is created in order to provide an introduction to the procedures in the wizard for4-dimensional data processing for BrainVoyager.

For this manual it is assumed that the reader is already familiar with the functionality in BrainVoyager. If,however, the meaning of certain parameters is unclear, please consult the documentation for BrainVoyagerwritten by Prof. Dr. Rainer Goebel. For the reader’s convenience, some of the information about pre-processing options in BrainVoyager has been included in this manual. For the up to date information onBrainVoyager’s procedures, please check the manual.See for example theonline BrainVoyager User’s Guide andRainer’s BV Blog.

http://brainvoyager.com/bv/doc/UsersGuide/BrainVoyagerUsersGuide.html

http://www.brainvoyager.com/bvresources/RainersBVBlog/index_bvblog.html

Chapter 1

Introduction

This document provides information about the use of the wizard to process multiple 4D data sets in Brain-Voyager.

1.1 Notes per wizard version

1.1.1 Notes on version 1.9 for BrainVoyager 20.6

This version contains FMR renaming, and the functionality to run the complete process from FMR creationuntil MTC file creation at once.

1.1.2 Notes on version 1.4 for BrainVoyager QX 2.4

In this version, the size of the VTC bounding box can be set numerically.


This version includes the since BrainVoyager QX 2.4.1 scriptable“GLM approach” for high-pass filtering.


This version includes VTC/VDW creation, VTC preprocessing and MTC creation for multiple files.


This version includes intra-session alignment for multiple files.


This version also contains creation of FMR and DMR projects.


Because the scripting module changed due to changes in the Qt Library 1 that is used in BrainVoyager, thebatch processing wizard had to be rewritten. All features in the 4d batch processing wizard will graduallyre-appear in this rewritten version. We have tried to include as many as possible options for the currentversion.

1The Qt library for C++ can be found at: http://www.qtsoftware.com/|

1

1.1.8 Notes on version 0.6

This version now includes the creation of DMR and VDW files. Also, it is possible to rename DICOM filesin multiple directories. This version is for BrainVoyager QX 1.10.4 (because of the new scripting commandto create VDW files).


Version 0.5 beta 2

Run all steps in one go The main change in the transition from v0.4 to v0.5 is that it is now possible torun all processing steps sequentially. The wizard will still write the text files with names of the producedfiles after each processing stage.This version is for BrainVoyagerQX 1.9.10. The script project can only be used in BrainVoyager QX 1.9 orhigher; in 1.9.10 there are more interpolation options than previous 1.9 versions. A wizard for an olderversion of BrainVoyager can be created on request. The function to create DMRs has been adopted andmodified by Pim Pullens and is now separately available (on the wiki).

Tabs for filenames Whenever it is possible to select parameters for list of files, the dialogs are using tabsnow, so that the size of the dialogs stays more decent. Please go through all of the tabs before proceedingto the next dialog.

Default directory When running the wizard, one can indicate a default directory. This saves a lot ofclicking when selecting files. Also, the resulting log files can all be found now in this default directory.

Intra-session alignment In the previous version of the wizard, it was assumed that the same parameterswould be applied for each FMR project. In the current version one can select a target FMR project if neededand indicate the target volume for each FMR project.

Version 0.5 beta 3

Set inter-slice time manually This option has been added in the beta 3 version (29.04.08). See section3.1.4.

Single image file loops This option has been added in the beta 3 version (29.04.08). See section 3.1.3.

Improvements and bug fixes The intra-session alignment procedure uses now the number of the targetfile in the list instead of the name of the target file, since the name is due to changes throughout the prepro-cessing procedure.The order of processing has been made similar to the order of initial manual file selection.The “linear trend removal only” option was applied after the full high-pass filtering option. This has beenchanged to performing linear trend removal only when this option is checked.


The functions can be started via the BrainVoyager QX “Scripts” menu. The functions can be used directly,without modification of the scripts in the script project. The main functions will give a text file with FMRproject names as output, that can be used for further preprocessing steps. The projects will be written tothe same directory as the source files. All main functions have the same structure:

1. select filenames or list with filenames

2. specify parameters or read parameters file (only motion correction and vtc creation)

3. perform the processing

2

4. write a list with processed filenames to disk

The help functions are provided to write a filenames or parameters file, to visually inspect the process-ing result and to convert between filter units. For an explanation of the processing steps, please consultthe BrainVoyager QX Getting Started Guide or the User’s Guide. These documents can be found viahttp://www.brainvoyager.com/. For information on scripting in BrainVoyager QX, please consult the Au-tomation and Scripting guide that is available on the wiki at http://support.brainvoyager.net/.This version is for BrainVoyagerQX 1.9. The script project is not backward compatible. A new function tocreate DMR files is included.Thanks to Koene van Dijk and Michael Capalbo for suggestions.


Because of QSA differences, there are now two versions of the batch processing project, one for 1.8 and onefor 1.9. In this version, it is now possible to provide filenames for the FMR project. Also, it is not necessaryany more to indicate the number of projects in advance.In this documentation, background information on the image processing options has been added (from thetopic “FMR Preprocessing”→ “Update” topic in the BrainVoyager 2000 User’s Guide at:http://www.brainvoyager.de/BV2000OnlineHelp/BrainVoyagerOnlineHelp.html). For anup-to-date explanation of the preprocessing techniques, see also Goebel et al (2006, [1]). Date: May 2007.


Date: December 2006.

3

1.2 Installation of the wizard

1.2.1 Location of the plugin and script files

The script batchprocessingwizard_vxx.js and the interface batchprocessingwizard_vxx.uishould be located in the folder:/(My) Documents/BVExtensions/Scripts/ (see figure 1.1).

Figure 1.1: Location for the wizard

When the script files are in the proper location, BrainVoyager can be started. The batchprocessingwizardwill appear in the BrainVoyager “Scripts” menu, but because it has a dialog, needs to be started via theScript Editor (Scripts > Edit and Run Scripts...) .

4

1.3 Acknowledgements

Thanks to Dany d’Souza, Christoph Oberthuer, Hedy Kober, Johnston, Jochen Seitz and Bettina Sorger fortheir suggestions for earlier versions of the wizard. Thanks to Henk Jansma for ideas on the logo of thebatch processing wizard.

5

Chapter 2

Launching the wizard and global options

2.1 Description of the functions

The batchprocessingwizard, materialized in batchprocessingwizard.js+*.ui, covers most process-ing steps that can be scripted in BrainVoyager, from project creation to mesh time course (MTC) creation;also, diffusion weighted image projects (DMR projects) can be created with this wizard.The functions can be used in one procedure, where the output filenames serve as input for the next step,or separately, when it is for example necessary to perform temporal high-pass filtering on a bunch of func-tional data files. The processing that can be performed is:

creating FMR/DMR projects: Input: files selected via dialog or specified in file list file.Output: project files <original name>.fmr/dmr andfilenames_for_batchprocessingwizard_stage_projectcreation_v19.txt.

mean intensity adjustment: Input: filenames from previous step or specified in file list file.Output: _MIA.fmr files; andfilenames_for_batchprocessingwizard_stage_meanintensityadjustment_v19.txt.

slice scan time correction: Input: filenames from previous step or specified in file list file.Output: Project1_SC*.fmr to ProjectN_SC*.fmr andfilenames_for_batchprocessingwizard_stage_slicescantimecorrection_v19.txt.

motion correction: Input: filenames from previous step or files selected via dialog or specified in file listfile.Output: *_3DMC.fmr files. Andfilenames_for_batchprocessingwizard_stage_motioncorrection_v19.txt.

temporal filtering: Input: filenames from previous step or files selected via dialog or specified in file listfile.Output: *_LTR_T*.fmr files andfilenames_for_batchprocessingwizard_stage_temporalfiltering_v19.txt.

spatial filtering: Input: filenames from previous step or files selected via dialog or specified in file list file.Output: *_SGS*.fmr files andfilenames_for_batchprocessingwizard_stage_spatialfiltering_v19.txt.

coregistration: Input: filenames from previous step or files selected via dialog or specified in file list fileand *.vmr.Output: *_IA.trf, *_FA.trf.

creation of VTC files: Input: files selected via dialog or specified in file list file.Output: ‘*.vtc’ files and‘filenames_for_batchprocessingwizard_stage_slicescantimecorrection_v19.txt’

6

creation of VDW files: Input: files selected via dialog or file list file.Output: ‘*.vdw’ files and‘filenames_for_batchprocessingwizard_stage_resulting_vdw_v19.txt’

preprocessing of VTC files: Input: Output: ‘*.vtc’ files andfilenames_for_batchprocessingwizard_stage_preprocessedvtcs_v19.txt.

creation of MTC files: Input: files selected via dialog or specified in file list file.Output: ‘*.mtc’ files and‘filenames_vtc_for_mtc_VTC.txt’ and/or‘filenames_for_batchprocessingwizard_stage_mtcfiles_v19.txt’

The files are written to the same directory as where the source files reside. If several steps are performedsequentially, the files only need to be selected once; for example, if one starts by creating projects and endswith creating MTC files, one only needs to specify the data that came from the scanner (for example *.dcm)and the supporting files (*.vmr, for coregistration and for creating the *.vtc file(s)).The text files with lists of produced files are written to the main directory that can be indicated after theproject parameters have been entered (see figure ??). In case a text file with the same name already exists,this might be overwritten. So in case there are important file list files, it might be helpful to transfer theseto another directory.

7

2.2 Global options

The wizard can be started via the BrainVoyager ‘Scripts’ menu (see figure 2.1). Select the menu option ‘Editand Run Scripts...’. Then, double-click batchprocessingwizard.js on the left hand side of the Script Editorthat appears, until the script is visible in the main window of the Script Editor. Then click on the ‘Run’button.

Figure 2.1: Launch the wizard via the BrainVoyager ‘Scripts’ menu

A main dialog will appear appear indicating all possible processing steps that can be performed at a time(see figure 2.2). The wizard will only accept parameters of options that have been selected on this globaloptions tab; all other values will be ignored. Files that belong to the selected target space and source filesneed to be selected. In case of intra-session alignment, the target *.fmr needs to be prepared before startingthe wizard.In addition to source files (scanner files, for example *.dcm, or *.dmr/*.fmr), and a target *.fmr in case ofintra-session alignment, the following files will be asked per target space:

Native/scanner space: Files to be selected are native files like *.dcm or *.PAR/*.REC; also, if coregistrationtakes place or VTC or VDW files are to be created, *.vmr file(s) will be asked for.

AC-PC space: If AC-PC space is selected, *.vmr, *_ACPC.vmr and *_ACPC.trf files will be asked for, ifVTC files are also to be created.

Talairach space: If Talairach space is selected, *.vmr, *_TAL.vmr, *_ACPC.trf and *.tal files will beasked for, if VTC files are also to be created.

MNI space: If MNI space is selected, *.vmr, *_MNI.vmr and *_a12.trf files will be asked for, if VTCfiles are also to be created.

8

Figure 2.2: Global options of the wizard

9

2.3 Processing order

2.3.1 Order of processing of data sets

The order of file selection determines the order of processing. So in case you think one data set should beprocessed before another one, follow the same temporal order as the wizard should follow later on.

2.3.2 Order of processing steps

When multiple data sets are selected, each step will be processed for all data sets before proceeding to thenext step. So first all FMR projects will be created before any slice scan time correction will occur. For theFMR preprocessing, the wizard assumes that slice scan time correction precedes any other preprocessingstep.The processing steps are applied in the following order:

1. Project creation

2. Mean intensity adjustment of *.fmr

3. Slice scan time correction *.fmr

4. Motion correction and intra-session alignment *.fmr

5. Temporal filtering of *.fmr

6. Spatial filtering of *.fmr

7. Registration between *.fmr and *.vmr

8. Creation of *.vtc or *.vdw files

9. Spatial filtering of *.vtc files

10. Temporal filtering of *.vtc files

11. Creation of *.mtc files

When you prefer another order, apply the steps individually.

2.4 Reports

2.4.1 BrainVoyager Log tab

The selected files and processing options are printed to the BrainVoyager Log tab. After selecting the text,a right-click provides the possibility to copy the text. This text can be pasted and saved in a simple text file.

10

Chapter 3

Creating FMR or DMR projects

3.1 Specifying parameters

3.1.1 Creating FMR projects

The options to create FMR projects are similar to the ones when creating a single FMR project via ‘File’→‘New’ in BrainVoyager, although the difference is that the headers of the files are not read automatically.This means that the acquisition parameters should be entered carefully; the parameters that are displayedin figure 3.1 are simply default parameters.

Figure 3.1: Options for creating FMR or DMR projects

3.1.2 Creating DMR projects

To create multiple DMR projects, can use here the same parameters as when creating DMR files via ‘File’→‘Create Project Wizard’; just like with creating DMRs via this batch processing wizard, headers of the filesare not read automatically, therefore please take care to enter the acquisition parameters that are applicableto your files. In creating DMR projects, one can use the following three filetypes: DICOM (*.dcm or *.IMA),

11

Philips (*.PAR and *.REC) or Analyze (*.hdr and *.img).The settings for creating the BrainVoyager exampleDTI dataset are displayed in figure 3.2.

Figure 3.2: Create example DMR project

For further processing of DMR files, please see the DTI Getting Started Guide by Pim Pullens. This guideis installed with BrainVoyager (can be found in the folder/Applications or Program Files/BrainVoyager/GettingStartedGuides/ or please see or sup-port sitehttp://support.brainvoyager.com/diffusion-weighted-imaging.html.

3.1.3 Storage order of the slices

Usually all slices of a volume acquired at a timepoint are stored before a new volume is stored: slice 1, 2 . . . nat timepoint 1, slice 1, 2 . . . n at timepoint 2, to slice 1, 2 . . . n at timepoint n. In some cases, all timepoints ofa slice are stored together before a new slice is stored; so in this case the slices are not stored volume-wise,but slice-time wise.

Figure 3.3: Set slice-wise storage order of slices (in contrast to volume-wise storage order)

If your functional MRI data are stored in the slice-time fashion, use the “slices× time” option to ensure thatthe data are read properly (see figure 3.3).

3.1.4 Set project parameters manually

If the interslice time is not set explicitly as a property, it will be calculated by BrainVoyager. The calculatedinterslice time is TR/number of slices. In the wizard it is possible to enter an interslice time that isdifferent from the calculated interslice time. This can be necessary in cases when the TR is longer than thetime of actual data acquisition, for example when the MRI experiment is combined with another modality

12

in the scanner.In the wizard, it is possible to alter the following parameters:

• TR

• inter slice time

• pixel resolution on x-axis

• pixel resolution on y-axis

• slice thickness

If the actual interslice time is different from the calculated interslice time, check the “Use new inter slicetime” box and enter the actual interslice time (see figure 3.4). Please take care that the new interslice timeis preserved throughout the processing (on some dialogs, BrainVoyager displays the calculated interslicetime instead of the interslice time that is saved in the FMR project).If the interslice time of your data is different from the calculated interslice time, use this option otherwisethe slice scan time correction might need be performed correctly. For the example diffusion weighted dataset, we have to reset the slice time and TR.

Figure 3.4: Set inter-slice time and TR manually

The DMR project will be saved with the same name as the raw data file (DICOM in this case, but with new*.dmr/*.dwi extension) and with the new parameter values.

13

Chapter 4

Mean intensity adjustment


For mean intensity adjustment, no parameters need to be specified.

4.2 Background: mean intensity adjustment in BrainVoyager

4.2.1 From User’s Guide of BrainVoyager QX 1.8

The Release Notes of BrainVoyager QX 1.8 state the following about Mean Intensity Adjustment:

Correction for global mean changes (from volume to volume) are usually not necessary.In the data from some scanners, however, we observed, quite substantial fluctuations (e.g. “spikes”), whichshould be removed during preprocessing (besides improving scanner performance). The previously exist-ing “Mean Intensity Adjustment” (MIA) option in the “FMR Data Preprocessing” dialog could be used forthis purpose, but this function worked at the slice level. This function has now been updated to work atthe volume level. During operation, the program now plots two curves, one showing the measured meanintensity of volumes over time and the other showing the mean level of each volume after correction (astraight line). Furthermore, a zero-mean predictor of the global fluctuations is automatically stored to diskallowing to add it as a confound predictor in the design matrix.If the MIA preprocessed FMR is used for further processing (e.g., VTC creation), adding of the MIA con-found predictor is not necessary. If one wants to perform the correction as part of the GLM, however, oneshould use the non-MIA FMR for further processing adding the MIA confound predictor to the designmatrix.

14

Chapter 5

Slice scan time correction


For slice scan time correction, only two parameters need to be indicated (see figure 5.1). These are the sliceorder and the interpolation method.

Figure 5.1: Options for slice scan time correction

In case of multiband data, the batchprocessingwizard will automatically recognise this by looking for theHasSliceTimeTable property; this property was introduced in BrainVoyager QX 2.8.2. In case a slicetime table has been detected, the wizard will use the script function CorrectSliceTimingUsingTimeTable()with cubic interpolation.

5.1.1 Slice order

The file names contain the characteristic indicating by SC that the FMR project was slice scan time cor-rected, and an additional marker to indicate the slice order. A is used for ascending, D for descending, I forinterleaved and 2 is added if there are interleaved in a special way (Siemens-specific).

• SCLA: ascending

• SCLAI: ascending interleaved

• SCLAI2: ascending interleaved 2

• SCLD: descending

• SCLDI: descending interleaved

• SCLDI2: descending interleaved 2

15

Concerning the interpolation method, *_SCL--.stc is an abbreviation of linear interpolation, *_SCC--.stcof cubic interpolation and *_SCS--.stc of SINC interpolation. Information about the meaning of slice or-der in BrainVoyager from the User’s Guide can be found in section 5.2.2.

5.1.2 Interpolation method

One can currently choose from 3 interpolation methods for slice scan time correction, which are linearinterpolation, cubic spline interpolation and (windowed) SINC.For a comparison of different interpolation methods, see the new section in the BrainVoyager User’s Guidewritten by prof. Goebel that has been included below (see section 5.2.1). For more information aboutinterpolation methods, see for example Meijering et al [2] or Thevenaz et al [3].

16

5.2 Background: slice scan time correction in BrainVoyager

5.2.1 From User’s Guide of BrainVoyager QX 1.9

An important step of preprocessing, especially for event-related designs, is the proper treatment of differ-ences in the time when individual slices are recorded. The problem of different slice scanning times stemsfrom the fact that a functional volume (e.g. whole brain) is usually not covered at once but with a seriesof successively measured 2D slices. For a functional volume of 30 slices and a volume TR of 3 seconds, forexample, the data of the last slice is measured almost 3 seconds later than the data of the first slice. Despitethe sluggishness of the hemodynamic response, an imprecise specification of time in the order of 3 secondswill lead to suboptimal statistical analysis, especially in event-related designs.One way to cope with slice scanning time differences would be to shift the expected BOLD time coursein time to compute proper statistical results. In this predictor-shifting approach, the reference time coursesfor a slice are shifted in time proportionally to the temporal difference in san time with respect to the ref-erence (e.g. first) slice. Another possibility is to shift the data of a slice in time to the same time point aswhen the reference slice was scanned. This changes the data in a way as if the whole volume would havebeen measured at the same moment in time. Note that in the latter case, the same predictors can be usedthroughout the volume, i.e. slice-specific shifts of the precictors are no longer necessary. For this reason, thelatter approach is normally used in fMRI data analysis, allowing to use the same predictors also after trans-forming the slice-based representation of the functional data (FMR/STC)to a 3D representation of the data(VTC) in an arbitrary (e.g. AC-PC or Talairach) space. It also allows to compare and integrate event-relatedresponses from different brain regions correctly with respect to temporal parameters such as onset latency.The predictor-shifting approach would require that each voxel would keep a label indicating the slice fromwhich it originates so that the correct shifted set of predictors could be used. Note that the data-shiftingapproach requires interpolation (resampling) of the slice time courses (see figure 5.2).

Figure 5.2: The data-shifting approach requires interpolation (resampling) of the slice time courses

17

Slice Scanning Order

In order to correct for different slice scan timings, the time series of individual slices are temporally “shifted”to match a reference time point, e.g. the first or middle slice of a functional volume. This temporal shiftdepends on the order in which the individual slices of a volume are scanned. Besides an ascending ordescending order, slices are often scanned interleaved, i.e. the odd slice numbers are recorded first fol-lowed by the even slice numbers. The figure above shows an example of ascending scanning (top) andinterleaved scanning order (bottom) for five slices. The time points of original scan times are indicated withgray rounded rectangles, while the shifted data are depicted with pink rectangles. The resampled data canbe treated as if all slices of one functional volume would have been scanned at the same time point.

Temporal Interpolation

As described above, correcting differences in slice timing using the data-shifting approach requires to re-sample the data at time points falling in-between measured data points. The values at the non-measuredtime points can be estimated by using measured data points ”in the neighborhood” i.e. from time pointsmeasured in close proximity. In BrainVoyager QX, three interpolation methods are available, linear, cubicspline and windowed sinc interpolation. The linear interpolation method uses only the neighbor on the leftand on the right side to calculate the value at the resampled point. It implements an average of these valuesweighted by the relative distance: xtnew = (1 − δ) ∗ xt − 1 ∗ δ ∗ xt. The linear interpolation method is fastto compute but has the disadvantage that it smoothes the data. Furthermore, the introduced smoothingdepends on the slice, i.e. it is negligible if points are resampled close to a measured point but smoothingis rather strong if points are resampled in the middle between two measured points (see below). The twoother interpolation schemes, cubic spline and since interpolation, avoid this problem by using more pointsin the neighborhood leading to a very accurate resampling.In the following, the different interpolation methods are compared using the first 50 volumes of the ”Ob-jects” sample data set (the first two volumes have been skipped, i.e. volumes 3 - 52 are used from theoriginal data). The sample data consists of 25 slices, which are scanned within 2 seconds (TR) in an inter-leaved ascending order.

Figure 5.3: The original time course of a voxel in red from slice 22

Figure 5.3 shows the original time course of a voxel in red from slice 22, which is the slice before the lastslice scanned within a volume. In this case, the data has to be shifted forward almost for one full TRand the figure shows that all interpolation methods are performing this shift very well without significantdeviations from each other. The curve from the linear interpolation method is shown in green, the cubicspline method in blue and the sinc method in yellow. Since the sinc curve is plotted last, the linear andcubic spline curve are hidden behind due to the similarity in the interpolation result.In figure5.4 the raw data (red) and the result from the three interpolation methods are shown for slice 23,which is the slice scanned prior to the middle slice. In this situation, the time course must be shifted abouthalf a TR, i.e. time points must be “recreated” in the middle between two measured time points. In this

18

Figure 5.4: The raw data (red) and the result from the three interpolation methods are shown for slice 23

situation, the cubic spline (blue) and sinc (yellow) interpolation method are performing very similar butthe linear interpolation method introduces visible smoothing. This is shown in more detail in the two sub-sequent figures.

Figure 5.5: Comparison of the cubic spline (red) and sinc (green) interpolation method: they produce iden-tical results (the green curve hides the red curve)

Figure 5.5 compares the cubic spline (red) and sinc (green) interpolation method, showing that they produceidentical results (the green curve hides the red curve). Figure 5.6 below compares the interpolation resultfrom the cubic spline (red) and the linear (green) interpolation method. Here it becomes clear that the linearinterpolation method smoothes the data as is apparent in weaker up and down deflections in the green ascompared to the red curve.The examples confirm the theoretically expected behavior of the interpolation methods demonstrating slice-dependent smoothing of the linear method. Therefore cubic spline or windowed sinc interpolation shouldbe used for slice scan time correction. Cubic spline interpolation is set as the default method in the Prepro-cessing dialog (see snapshot of the right upper part of the Preprocessing dialog in figure 5.7 below).

5.2.2 From User’s Guide for BrainVoyager 2000

The slices comprising one functional volume are scanned sequentially, i.e. at different moments in time.For functional analysis, i.e. in event-related designs, a whole functional volume is, however, treated asone data point, as if all slices were measured at the same time. To make this treatment of the data valid(i.e. for interpreting time the same way across a functional volume), the sequentially scanned slices haveto be interpolated in time. For a correct temporal interpolation of the slices, the slice scanning order has

19

Figure 5.6: Comparison of the cubic spline (red) and linear (green) interpolation method: the linear inter-polation method smoothes the data

Figure 5.7: Slice scan time correction: in right upper part of the Preprocessing dialog

20

to be known, which can be ascending or descending and either interleaved or not. The scanning order ofthe slices together with the TR and inter slice time can then be used to perform an appropriate temporalinterpolation. You can either use sinc (default) or linear interpolation by selecting the Linear or Sinc option,respectively, in the Interpolation field within the Slice scan time correction field (see snapshot below).

Figure 5.8: Slice scan time correction: in right upper part of the Preprocessing dialog (BV2000)

The specification of the slice order has to be performed with care because different manufacturers define”ascending” differently, i.e. either from the top of the head to the bottom or vice versa. In addition, Brain-Voyager allows to reorder the slices (Reverse order button in the Edit FMR File Specifications dialog). Toavoid any misconception, BrainVoyager now provides a Slice Scanning Order dialog to aid in the correctslice order specification. To invoke this dialog, click the Options... button in the Slice scan time correctionfield. The slice order dialog separates two pieces of information: the display order and the scan order of theslices. The display order is shown as a sequence of thumbnail images and is identical to the order shown inthe FMR project window (see snapshot below).

Figure 5.9: Slice order: descending

The correct scan order of the slices has to be specified below the thumbnails with respect to the displayorder. In the example data set shown above, the images are displayed from the top of the brain (left upperslice shown) to the bottom of the brain (right lower slice shown). If the slices would have been scannedalso from the top to the bottom, the correct selection would be the Ascending option in the Specify slicescanning order field. In this example, the slices were, however, scanned from the bottom to the top sothat the correct selection is the Descending option (see snaphsoth above). The labels below the thumbnails(”Scan: 1”, ”Scan: 2” etc.) reflect now the real scan order, i.e. the bottom slice was scanned first. Note that

21

this specification is correct irrespective of whether the bottom-to-top scan order was labelled ”ascending”or ”descending” by the manufacturer. What counts is the specification of the scan order with respect tothe display order. Scanning is often performed in ”interleaved” mode in which first the odd (slice 1, 3,5...) and even (slice 2, 4, 6...) numbered slices are scanned. The Interleaved option can be combined withthe Ascending and Descending option. As an example, ”ascending, interleaved” has been specified in thesnapshot below. With respect to the display order, the ”Scan: <x>” labels below the thumbnails showaccordingly that the first slice displayed was scanned first, the third slice displayed was scanned second,the second slice displayed was acquired as scan 11 and so on.

Figure 5.10: Slice order: ascending

An option Interleaved 2 is additionally available for Siemens scanners with the Numaris 4 software. In thislatest software, ”interleaved” is defined differently with regard to an odd or even number of slices. Brain-Voyager takes care about this particular definition when using the ”interleaved 2” option.The file name segment reflecting slice scan time correction is _SC[SL][A—D][I—I2]—. The _SC part isalways present to denote ”slice scan time correction”. This is followed by either letter ”S” denoting ”sincinterpolation” or letter ”L” denoting ”linear interpolation”. This is then followed by either letter ”A” de-noting ”ascending” or letter ”D” denoting ”descending”. Optionally this is followed by the letter ”I” orthe letters ”I2” denoting ”interleaved” scanning. As an example, the string _SCSAI in an FMR file nametells us that slice scan time correction had been performed using sinc interpolation and with an ascendinginterleaved scan order specification.

22

Chapter 6

Performing motion correction andintra-session alignment


For motion correction, the global parameters can be provided via the ‘3D Motion correction and intra-session alignment’ section on the preprocessing tab of the wizard (see figure 6.1). These are the sameparameters as can be entered via the “Preprocessing” dialog of the BrainVoyager menu.

Figure 6.1: Options for motion correction

Interpolation options are

• trilinear motion estimation and correction

• trilinear motion estimation and SINC motion correction

• SINC motion estimation and correction

The default setting in BrainVoyager is trilinear estimation and SINC motion correction (see also figure 6.2).

Figure 6.2: Interpolation for motion correction

The difference between the motion correction log (*_3DMC.log) and extended log (*_3DMC_verbose.log)is that for the default log, the rotation and translation parameters per volume are given in millimeters anddegrees:-> volume: 2 n_its: 3dx: 0.0000 mm dy: 0.0000 mm dz: 0.0000 mmrx: 0.0000 degs ry: 0.0000 degs rz: 0.0000 degs

23

while in the extended log file the gradients for translations and rotations, and the error and λ are providedper iteration for each volume:

volume: 2 iteration: 1 Error: 54.015769 Lambda: 1.000000e-04GradientX: -0.393465 TranslX: 0.001868GradientY: 0.835942 TranslY: 0.005183GradientZ: 30.727486 TranslZ: -0.000186GradRotX: 1.723598 RotX: -0.007159GradRotY: 0.364815 RotY: 0.001076GradRotZ: -0.006980 RotZ: -0.001683

The target volume refers to the volume in the run that the other volumes are aligned to. In motion correctiononly this refers to a volume in the same run; in motion correction and intra-session alignment the targetvolume refers to the volume in the target run.The generation of movies can be set to on (checked) or off (unchecked).The intra-session alignment option is available from wizard version 1.0. To apply intra-session alignment,processed FMRs must be available to serve as target FMRs. The filenames of the target FMRs can be selectedafter selecting each source file via the file dialog, or placed in a list after the list of source filenames in a textfile (see section 13).

6.2 Background

From the support site:“There are different options available in BrainVoyager QX allowing to adapt the method to your require-ments. First of all you can choose between three different interpolation methods: trilinear, sinc, or a com-bination of trilinear for detection of motion and sinc for correction. The method used is also indicated inthe resulting filename following the _3DMC with T for trilinear, S for sinc and TS for a combination of both.While the computation cost of sinc interpolation is very high, trilinear interpolation slightly smoothes thedata spatially. Therefore, it is recommended to use the “Trilinear/sinc interpolation” option (default) toavoid the problem of inducing unwanted blurring effects in the data while preserving a reasonable com-putation time. Furthermore, it has been shown that it is not necessary to use all voxel values for estimationof the movement parameters. Hence, BVQX uses by default a reduced number of voxels which can bechanged by deselecting the “Reduced data” option.. . .You can choose the reference volume for the dataset (default is the first volume) or even another FMR foralignment (of course acquired in the same session!). The last mentioned option would make sense, forexample if another functional data set was acquired closer to the anatomical data (to improve the coregis-tration of both data sets later on). However, it is important to stress the point that this option is not meantfor aligning different sessions, since it cannot easily correct big misalignments. By default the T1 saturatedfirst volume fmr is also aligned to the reference volume. The reason for that is that it is often recommendedto use the first volume fmr for the coregistration of the functional and anatomical data set, since the firstvolume is characterized by a more pronounced contrast which improves the alignment of the data.Note: Although the _firstvol.stc is changed by the motion correction procedure (when the defaultvalues are kept) it is not apparent in the file name!. . .BrainVoyager masks out noise voxels by default using only voxels with an intensity above 100, when thisvalue is set to 0 also voxels of the background are used for the motion correction process.In the field “Parameter estimation for a volume” you can specify the maximum number of iterations usedfor each volume. This value is set to 100 which is very conservative and usually not necessary to estimatethe motion parameters.. . . ”

24

http://support.brainvoyager.com/functional-analysis-preparation/27-pre-processing/273-motion-correction.html

Chapter 7

Temporal filtering


High-pass and low-pass temporal filter parameters can be entered via the ‘Temporal filtering’ section onthe “Preprocessing” tab of the wizard (see figure 7.1). The filter value can be provided in spectral units,either cycles per second (Hertz: “Hz”) or cycles per time course (“cycles”).

Figure 7.1: Parameters for temporal filtering

The default in BrainVoyager is High pass GLM Fourier with 2 sines/cosines.When choosing to filter in Fourier space (Frequency filter), one could use for example a cut-off value of128 seconds. The filter will remove the lower frequencies that repeat over a longer period than the cut-off.“The period is the duration of time of one cycle in a repeating event, so the period is the reciprocal of thefrequency.” (Wikipedia). The frequency f = s−1 in Hertz is the reciprocal of the period T in s: f = 1

T .Hence a cut-off value of 128 seconds corresponds to a frequency of 1/128 = 0.0078125Hz.

25

https://en.wikipedia.org/wiki/Frequency

7.2 Background: temporal filtering in BrainVoyager

The information below is borrowed from the BrainVoyager 2000 User’s Guide, just to provide general in-formation about the filter units.

7.2.1 Introduction

Temporal filtering. Temporal filtering consists of two main functions: removal of drifts and temporal gaussiansmoothing.

7.2.2 High pass filtering

The removal of drifts is very important and is highly recommended whereas temporal smoothing is lessimportant and not checked as default.The drifts observed in voxel time series are often linear but can be also nonlinear (see time course below).

Figure 7.2: The drifts observed in voxel time series are often linear but can be also nonlinear

The minimum you should do is a linear trend removal and, thus, the Linear trend removal option is checkedas default. To remove also nonlinear drifts, a temporal high pass filter is recommended. This filter convertsthe time series at each voxel in the frequency domain (using a Fast Fourier Transform or FFT) and thenremoves low frequencies (= high pass). The frequency domain representation is then converted back in thetime domain (inverse FFT), which looks then as before except that low frequency drifts are no longer visible.Since this filter is desirable in most cases, the option High pass filter: is checked as default. Prior to a highpass filter, linear trend removal is always performed because this improves the result of high pass filteringin case of a large linear drift component. You can specify which (low) frequencies should be removed eitheras cycles in time course, cycles/point or Hz. The unit selected as default is “cycles in time course” and thepreset value is 3. This value needs not to be changed except in the case that your paradigm uses only a few(i.e. 3 or less) blocks or trials. In this case, you should reduce the value to 2 or 1 or turn high pass filtering off.Note, however, that it is highly recommended to use (much) more than 3 blocks or trials per run in order tobe able to separate fluctuations due to true activation effects from nonlinear drifts due to scanner and phys-iological noise. If you chose to enter the high pass filter in cycles per point or Hz, the program converts thecycles in time course units for you. Note that using the unit Hz requires a correct specification of the TR value.

7.2.3 Gaussian temporal smoothing

Besides removal of drifts, you can also apply a gaussian temporal smoothing filter. Since temporal gaussiansmoothing blurs timing information across neighboring data points, it is not recommend as default. Tem-poral smoothing improves, however, the signal-to-noise ratio by removing high frequency fluctuations. Ifyou want to apply temporal smoothing, check the Gaussian option in the Temporal smoothing parameters

26

field. The width of the kernel can now be specified in seconds. Note that the specification in seconds is onlycorrect if the TR value has been specified correctly.The kernel width is specified in the Gaussian - FWHM: text box with the default value of 2.8 seconds. If youwant to specify the width of the kernel in units of data points (TR’s), check the data points option insteadof the secs option.The file name segment reflecting temporal smoothing consists of up to three parts. If linear trend removaland/or high pass filtering is performed, the string _LTR (linear trend removal) is appended first. If a highpass filter is applied, the substring _THP<filter><unit> is appended. This is followed by the filter valueand the unit of that value, which is either “c” (cycles in time course), “cp” (cycles per point) or “Hz”. Iftemporal smoothing is performed, the substring _TDTS<FWHM><unit> is finally appended. The substring“TD” is short for “time domain”, the substring “TS” is short for “temporal smoothing”. This substring isfollowed by the width of the gaussian kernel (<FWHM>) followed by the unit of that value, which is either“s” (seconds) or “dp” (data points). As an example, the string _TDTS2.8s in an FMR file name tells us thattemporal smoothing had been performed in the time domain with a gaussian kernel of 2.8 seconds FWHM.

27

Chapter 8

Spatial filtering


Spatial low-pass (Gaussian) filter parameters can be entered via the ‘Spatial smoothing’ section on the“Preprocessing” tab of the wizard dialog (see figure 8.1). The width (FWHM, see below) of the filter kernelis specified in spatial units - either millimeters or pixels.

Figure 8.1: Parameters for spatial filtering

28

8.2 Background: spatial smoothing in BrainVoyager

The information below is from the BrainVoyager 2000 User’s Guide, just to provide general informationabout the filter units.

8.2.1 Introduction

Spatial smoothing. Spatial smoothing uses a 3D gaussian kernel. The width of the kernel can now bespecified in millimeters. Note that the millimeter specification is only correct if the “voxel resolution”has been specified correctly (or more conveniently if it was extracted from the file header during projectcreation). The kernel width is specified in the Gaussian filter - FWHM: text box (FWHM = full width at halfmaximum). If you want to specify the width of the kernel in pixels, check the Pixel option in the Spatialsmoothing parameters field. If you want to run gaussian smoothing only within the slices, check the 2Doption instead of the 3D default option. Spatial smoothing is executed in the space domain (image space)as default, which is fast for small kernels. For large kernels (more than 15 mm), smoothing in the frequencydomain is often faster. If you want to use frequency domain gaussian smoothing, check the Freq. domainoption.To exploit the full spatial resolution of fMRI, we do not recommend to spatially smooth the data for singlesubject analysis. This is the reason why the Spatial data smoothing option is turned off as default in thePreprocessing option field. If you want to smooth your data for single subject analysis, we recommend touse a small kernel of about “4 mm”. For multi-subject analysis, we recommend a kernel between “8” and“12” mm. We suggest to apply spatial smoothing for multi-subject analysis after having first run single-subject analyses with no or modest spatial smoothing. The recommended moment for multi-subject spatialsmoothing is at the stage where you want to run multi-subject GLM analyses. At this stage, you will haveone or more VTC files per subject. To run spatial smoothing at this stage, you must use the version forVTC files, which is available in the 3D Data Preprocessing dialog (nb: VTC smoothing is now available viascripting and will be placed in a wizard soon). The file name segment reflecting spatial smoothing is

*_[SD|FD][3D|2D]SS<FWHM><unit>

. The substring “SD” is short for “space domain”, the alternative string “FD” is short for “frequency do-main”. The next substring is either “3D” or “2D” followed by “SS”, which is short for “spatial smoothing”.This substring is followed by the width of the gaussian kernel (<FWHM>) followed finally by the unit of thatvalue, which is either “mm” (millimeter) or “px” (pixels). As an example, the string *_SD3DSS4.00mmin an FMR file name tells us that 3D spatial smoothing had been performed in the space domain with agaussian kernel of 4 mm FWHM.

29

Chapter 9

Coregistration


In BrainVoyager, image registration is required to create a 3D functional image (*.vtc) that can be overlaidon an anatomical image (*.vmr). In BrainVoyager, the registration consists of two steps; the first step, thatis also made available via scripting, is called initial alignment, which is a mathematical mapping fromfunctional to anatomical image. The second step is called fine alignment. Via scripting, one can select the“normalised gradient fields (NGF)” approach for fine alignment or “boundary-based registration (BBR)”,which involves finding the cortical sheet in the volume from a surface file. These options are also availablein the batchprocessingwizard (see figure 9.1).

Figure 9.1: Options for coregistration

The coregistration step will result in two transformation files: *_IA.trf and FA_.trf. These are the esti-mated parameters for transforming the functional image.Because coregistration is now available as a scripting function, this feature has been included in the batch-processingwizard since 2017 (batchprocessingwizard v1.9).

9.2 Background: coregistration (and spatial normalisation) in Brain-Voyager

FMR-VMR registration in BrainVoyager results in transformation files that can be used to create a 3D func-tional file (*.vtc), which is either in native space or spatially “normalised” to a standard space, like Talairachor MNI space.For the normalisation to happen, one should also provide the batchprocessingwizard with files that bringthe anatomical image into the standard Talairach or MNI space, for example *_ACPC.trf and *.tal for

30

a 3D functional image (*.vtc) in Talairach space or *_a12.trf to obtain a *.vtc in MNI space. These needto have been made in advance. The surface file (*_RECOSM.srf) will be created during the BBR and so thename does not need to be provided to the batchprocessingwizard, except in the case the option is selectedto create functional files on the surface (*.mtc).

Figure 9.2: Options in BrainVoyager coregistration

31

Chapter 10

Creating VTC/VDW files


Figure 10.1: Parameters for creating VTC/VDW files

10.1.1 File type for normalisation

One can choose from either normalising functional data (based on *.fmr/*.stc files) or diffusion data (basedon *.dmr/*.dwi files). For both filetypes, initial alignment transformation files (*_IA.trf), fine alignmenttransformation files (*_FA.trf) will also be asked, as well as AC-PC transformation files (if applicable)and Talairach landmark files (*.tal)(if applicable) or MNI normalisation files (*_a12.trf).

10.1.2 Parameters for VTC/VDW creation

The data type of the VTC or VDW files can be float (32-bit) or integer (16-bit). The float data type is nowdefault in BrainVoyager.

32

10.1.3 Target space

The target space of the VTCs or VDWs can be native, AC-PC space, Talairach or MNI. In case Talairach isselected, one can choose to include the cerebellum by checking the ‘extended with cerebellum’ box.

10.1.4 Resolution

The resolution of a VTC or VDW file that is created via scripting can be 1, 2 or 3 mm (isotropic). In Brain-Voyager 2.4 it is also possible to create VTCs in higher resolution via the BrainVoyager user interface.

10.1.5 Interpolation type

The interpolation type can be nearest neighbour, linear or SINC.

10.1.6 Bounding box determination

There are three choices to determine the size of the bounding box.

Use default: This will use an intensity threshold ’100’ for non-Talairach space and for data in Talairachspace the default coordinates (x = 59 - 197, y = 57 - 231, z = 52 - 172).

Intensity threshold: This can be used for non-Talairach data. The default value for the threshold is ‘100’.

Indication of coordinates w.r.t. anatomy (*.vmr): This can be used in all reference spaces (native, AC-PC,Talairach, MNI). With this method, all VTCs will obtain the same size of bounding box.

33

Chapter 11

Preprocessing VTC files

From the BrainVoyager User’s Guide: “While motion correction and slice scan time correction can only beperformed for FMR projects, some preprocessing steps are also available for VTC projects that are createdafter spatial normalization, i.e. after creation of functional volume time courses in a known 3D space,such as native (space of a measured VMR data set), ACPC or Talairach space. Since VTC preprocessingincludes 3D spatial smoothing and temporal filtering/smoothing, it is not necessary to apply these stepsto the original FMR space which would require to repeat the (rather slow) VTC creation process. To retainflexibility and to minimize work load, we thus recommend to:

• Perform motion correction, slice scan time correction, high pass filtering (and MIA) in FMR space, butno spatial and temporal smoothing.

• Create VTC pendant for each FMR data set. Add smoothing as desired, e.g. little (4mm FWHM)or no spatial smoothing for single-subject analyses and cortex-based group analyses, and modestsmoothing (8mm FWHM) for Talairach space group analyses.”


11.1.1 VTC spatial smooting

One can select to smooth with the FWHM width of kernel in millimeters or voxels.

11.1.2 VTC temporal high-pass filtering

One can select a cut-off value specified in Hertz (= cycles per second) or cycles (= number of cycles in timecourse of particular file).

11.1.3 VTC temporal low-pass filtering (smoothing)

One can select to smooth with the FWHM width of kernel in seconds or data points.

34

Figure 11.1: Parameters for VTC preprocessing

35

Chapter 12

Creating MTC files


Figure 12.1: Parameters for creating MTC files

For creating MTC files, the only parameters that need to be specified are the sampling parameters in mil-limeters (see figure 12.1). The default in BrainVoyager is -1 to 3mm.

12.2 Background

From BrainVoyager’s User Guide:“If the option “Sample volume data exactly along mesh vertices” is selected, the mesh vertices are projectedinto the VTC data and the time series are read at the respective vertex positions. Since the vertex coordinatesare real values, they usually look for data at positions that fall in between the voxel grid; the values arethen determined by trilinear interpolation. If the option “Integrate data in depth along vertex normals” isselected (default), the value is determined by sampling and integrating several values along the directionof the vertex. The sampling step size is specified by the “Step value” (default: 1.0 mm) and the start valueis specified with the “From” value (default: -1.0). With these values the first sampling point is 1 mm awayfrom a vertex in the direction of white matter; the next value is 1.0 mm along the vertex normal in thedirection of grey matter, i.e. at the vertex position (-1.0 + 1.0 = 0.0). More values might be sampled until the

36

“To” value (default: 3.0) has been reached. The values sampled along the normal of a vertex are averagedresulting in the value attached to the vertex for one time point. The same sampling procedure is performedfor all time points in the VTC file producing the full vertex time series. After completing the sampling forall vertices, the mesh time course data has been created ...”

37

Chapter 13

File selection

Figure 13.1: Choice for file selection method

File selection is possible either via a file dialog or via a text file (figure 13.1) on the “Source file” tab. Figure13.2 indicates which files are required for each step.The batchprocessingwizard cannot cope with spaces in pathnames and filenames... So if the files are locatedin a folder “/Functional data/”, the wizard will stop. If the wizard is still useful in the next version ofBrainVoyager, hopefully this can be improved.

38

Figure 13.2: Required files per step; green v: always required for that step; yellow o: depends on previouschoices and selected target space

39

13.0.1 Selecting files manually

The procedure to select files manually is depicted in figure 13.3.First, select the option “Select files via filedialog”.Step two is to set the number of (functional/diffusion weighted) projects; for example, for 10 subjects witheach 3 runs, the number of projects to process is 30.Step 3 is optional, and only relevant for creating *.fmr or *.dmr projects. The default filename that thewizard uses, is the base part of the native file plus the *.fmr or *.dmr extension. For a different filename,please check the box “Edit future filenames”.Once ready, the files can be selected by clicking the button “Get input filenames...” (step 4). The files areselected per project; for example, if there are 3 projects to be created with FMR creation, preprocessing,and VTC creation, first the wizard asks for *.dcm file 1, then for the VMR 1 in native space, then for the*_ACPC.trf, *_a12.trf and/or *.tal files if the VTC is not in native space, and then for the VMR 1in target space (*_ACPC.vmr, *_MNI.vmr, *_TAL.vmr); this procedure is repeated for the second projectand then all files for the third project.Step 5: if the box “Edit future filenames” is checked, the name of the future *.dmr/*.fmr files have to beindicated on the “Rename” tab , including the extension (*.fmr for functional data or *.dmr for diffusionweighted data, see figure 13.4) after selecting all files via the “Get input filenames...” button (step 4).

Figure 13.3: Procedure to select files manually

Figure 13.4: Rename *.fmr or *.dmr data

The files will be asked per subject, from file creation, preprocessing, coregistration, normalisation, prepro-cessing to MTC file creation. See figure 13.2.

40

In case the option “Motion correction and intra-session alignment” has been chosen, after each file the tar-get FMR can be selected via the file dialog. Even if the source files are in other format, for example DICOM,the target filenames should be from FMR projects, which means one should first process the runs that onlyneed to be motion corrected and will serve as target, then the runs that need to be aligned to other runs.

13.0.2 Selection of files via a file list

The procedure of using a batchfile is depicted in figure 13.5.

Figure 13.5: Procedure to use batchfile

When the text file has been prepared (step 1) and all (pre-)processing parameters have been entered viathe main dialog, select “Read filenames from textfile” (step 2). The textfile containing the filenames for theprocessing can be selected via the “Get input filenames” button (step 3). A file dialog will appear once, toselect the batchfile (*.txt).The textfile has a fixed format since batchprocessingwizard v1.9 (2017). It always consists of twelve entriesper project. It starts with a number that indicates the number of projects. In case a file is not required,please write ‘none’. Anything can be written before the colon, but it is important to write a ‘:’ on each lineto separate the label, left, from the actual filename on the right hand side. The projects are separated by anempty line. The filenames should contain the full file path. An example to create a FMR project, coregister,create VTC files in native space and preprocess VTC files is shown below (please note that the commentsbetween parentheses should be removed before use):

1

1 scanner: /Volumes/DATA/testcases/wizard/A1/S1/S1-0002-0001-0001.dcm (functional file)2 fmr: none (this file will be created by the batchprocessingwizard)3 fmr ISA: none (apparently no intra-session alignment, otherwise provide an *.fmr)4 vmr for coreg: /Volumes/DATA/testcases/anatomy_IIHC.vmr (anatomical file)5 IA.trf: none (this file will be created)6 FA.trf: none (this file will be created)7 ACPC.trf: none (not provided here, so the *.vtc cannot be in AC-PC, MNI or Talairach space)

41

8 tal: none (not provided here, so the *.vtc cannot be in Talairach space)9 vtc: none (this will be created by the batchprocessingwizard)10 srf: none (this is not provided here, so no *.mtcs can be created)11 vmr final space: /Volumes/DATA/testcases/anatomy_IIHC.vmr (this is for a *.vtc in native space)12 mtc: none

Please note that the ‘vmr for coregistration’ (index 4) is only required for coregistration purposes, but the‘VMR in target space’ is required for the steps ‘VTC/VDW creation’, ‘VTC preprocessing’ and ‘MTC cre-ation’. If the analysis takes place in native space, the same VMR file name can be selected (or in case of textfile: placed in index 4 and index 11).

Figure 13.6: Example of batch file to create 2 FMR files on Mac OS X

The file selection order determines the order of processing of the data sets (see section 2.3).

42

For your information, a flow diagram of the functions in the batchprocessingwizard has been included (seefigure 13.7).

Figure 13.7: Different options for the wizard

43

Bibliography

[1] R. Goebel, F. Esposito, and E. Formisano. Analysis of fiac data with brainvoyager qx: From single-subject to cortically aligned group glm analysis and self-organizing group ica. Human Brain Mapping,27(5):392–401, 2006.

[2] E.H.W. Meijering, W.J. Niessen, and M.A. Viergever. Quantitative evaluation of convolution-basedmethods for medical image interpolation. Medical Image Analysis, 5(2):111–126, 2001.

[3] P. Thevenaz, T. Blu, and M. Unser. Interpolation revisited. IEEE Transactions on Medical Imaging,19(7):739–758, 2000.

44