OpenNFT: a real-time fMRI neurofeedback software

I am actively contributing to a project by Yury Koush (MRRC, Yale University) together with John Ashburner and Peter Zeidman (Functional Imaging Laboratory, University College London), Frank Scharnowski (University of Zurich), Dimitri Van De Ville (EPFL Geneva), Evgeny Prilepin, Sergei Bibikov and Artem Nikonorov (Samara State Aerospace University).

OpenNFT is an open-source framework for neurofeedback training based on real-time fMRI.

Check out our project website for additional details and to download the software:

sweetView: a simple, quick, and powerful viewer for MRI images and SPM results

We all love SPM. However, creating really nice figures for your talk or publication typically involves quite a lot of manual labor and post processing. And then, just when you are done and send the figures to your co-authors, you are informed that you need to exclude one subject from group analysis because their drug screening was positive (or negative – depending on the study).

Things like this happen and this was my motivation to create sweetView a simple and powerful viewer for MRI images and SPM results that allows you to quickly create triplanar or mosaic overlays of your SPM results. The core features include fast selection of images, easily customizable overlays for masks or SPMs, adding (anatomical) labels and saving the slice selection and multiple cursor positions, so you can easily reproduce your original figure design.

The software has been designed for people who use Matlab and SPM12.


Development roadmap:

  • sweetView v0.4. User experience. Rewriting user interface backend code, easier interface for global/local view settings, color picker, multiple windows (May 2017).
  • sweetView v0.6. Masking. Create masks, brain atlas integration, smart functional masks (July 2017)
  • sweetView v0.8. Time. 4D NIFTIs, time series, animations (Sept. 2017)
  • sweetView v1.0. Major release. Dissemination and release (Oct. 2017)



I created this tool a couple of years ago to automatically create larger model spaces based on a single template file. Sometimes there is uncertainty about the presence or absence of a connection in your model. In DCM you account for that by manually creating two models where this connections is either turned on or off and compare there respective model evidences. It is easy to see that this manual approach is tedious for larger model spaces, when there is uncertainty about many connections.

sweetDCMvariate Create DCM models within a given model space based on a given template file. Template files should be created with the official SPM functions or compatible implementation. This tool will then create all variations of connectivity priors as defined in the template model.

Download file: sweetDCMvariate.m
Need some help? Found a bug? Please contact me.