This Matlab (with some bash scripting) toolbox consists of three related components for designing, testing, and implementing real-time fMRI neuroimaging experiments.

Component 1. realTime Real-time fMRI data prep and processing.

Component 2. stimulusPresentation Stimulus presentation with psychtoolbox.

Component 3. qpfMRI A set of tools for using Quest+ with the temporal fitting engine in designing Bayesian staircase procedures.

The best way to install and use the neurofeedback tools is to install toolboxToolbox on Matlab.

1. Git clone into '/Documents/MATLAB/projects/'

2. If using with Toolbox Toolbox, then tbUseProject('neurofeedback')

3. Configure /config/neurofeedbackLocalHookTemplate.m and place in /MATLAB/localHookFolder

4. You will also need to install FSL.

Recommended directory structure:

• rtQuest
  • KastnerParcels
  • subject1
  • subject2
    • raw
    • processed
      • run1
      • run2

Top level is project-specific. (rtQuest)

Two types of directories within the main directory.

A templates directory in which you should place any ROIs that are in standard space. (KastnerParcels)

A set of subject-specific directories (the name of this directory should be able to be referred to as a string). Within that directory are two main directories: raw and processed

raw contains the MPRAGE and any sbrefs you want to use.

processed will contain all files generated through registration and neurofeedback. The scripts will put run-specific data in run-specific directories nested within processed

Before you get to the scanner (or, possibly very quickly while you are at the scanner), we need to set up the ROI.

Acquire an ROI. Any ROI will work.

realTime/brainProcessing has three scripts useful for preprocessing.

brain_extraction.sh performs bet brain extraction on the MPRAGE and the two ROIs.

makeMaskFromRetino.m creates a retinotopic mask from functional data. Unless you collected this specifically for your subject as part of a separate scan, it is unlikely that you can use it.

register_ROI_to_APandPA.sh will register the ROI (the way it's set up now will take a Kastner V1 parcel and register it to a scout EPI or an SBREF).

At the end, you should have an ROI registered to EPI space that can be used in runNeurofeedback.m. Specifically runNeurofeedback.m will call registerToFirstDICOM.m.

To be completed. This section will describe how to use runNeurofeedback.m.

If your system is set up correctly, you should be able to copy and paste the first set of example code from runNeurofeedback. Provided that you have raw DICOMs and set up your directories as described in that script (and getPaths), you should be able to copy DICOMs from the main directory into the "scanner directory" and see a live plot of the mean of the ROI you provided.