Fastball toolbox - New Psychopy based toolbox for steady-state oddball experiments

Hi,
I don’t know if this is appropriate place to post this, hopefully it is! We have developed a Psychopy and Matlab toolbox for steady-state EEG experiments and wanted to share it with other Psychopy users.

The Fastball toolbox allows you to measure the discrimination between categories of visual stimuli using EEG. The Psychopy scripts will build your experiment and present your stimuli using simple point-and-click interfaces, the Matlab code will analyse the EEG data recorded. Both scripts have many easily modifiable parameters and we have written a comprehensive wiki to guide users. We have deliberately designed the toolbox to be easily useable with only the most basic knowledge of Matlab or Psychopy.

Background
Based on a technique originally developed by Bruno Rossion (lab page here) the toolbox uses frequency tagging to measure the Mismatch Negativity Response (MMN). We recently successfully adapted the technique to measure semantic discrimination (paper here). We believe the great strengths of the technique are:

• Sensitivity – responses are consistently detectable in single subjects
• Speed – responses are stable and measurable in single subjects in 1-2mins of EEG recording
• Adaptability – Bruno Rossion’s lab has extensively developed the approach in face processing, we have adapted it to measure visual semantic discrimination and short term memory. It has also been used to measure discrimination between words and non-words. We will also be developing a version of the Psychopy scripts to present auditory stimuli.
• Clinical potential – the subject does not need to provide any behavioural or verbal response, making it ideal for objectively measuring cognitive abilities in populations with communication difficulties or cognitive impairment.

Where to download
The toolbox is available to download from here:

Dissemination
We would appreciate any help in the dissemination of the toolbox. Please feel free to share on twitter, researchgate etc. You can follow me on Twitter @GStothart, where I will be posting news and updates about the toolbox.

Open source
This code is free to be adapted, improved and generally messed around with. Just use the Github protocols for uploading and sharing new versions, detailed in the wiki.

We hope that you and your students find this toolbox useful, feel free to email me with any questions or suggestions.
Many thanks

George Stothart

8 Likes

This is super cool - definitely a great idea to make it easy to produce these stimuli.

I have a package that is helpful in analysing these: www.janfreyberg.com/ssvepy Might be worth combining forces for the python side of your toolbox so that people can just pip install it?

4 Likes

Thanks Jan! Glad someone thinks it might be helpful :slight_smile:
I like the idea of bundling your analysis code, but we’d need to make it clear that you need MNE pre-processed data for it to work, our Matlab script will work with anything in plain old text file format.
I’ll drop you an email as it looks like we have similar approaches with SSVEP analysis, be useful to compare notes!
Cheers
George

Sounds good, I’ll send you my email in a DM.

Yes, good point - I have a different approach and definitely don’t want to provide a start-to-finish pipeline. Maybe the best approach would be to make a separate pypi package for your package, but always install it with ssvepy and provide a unified interface that way - but we can talk about that!