Overview

While statistical maps of neuroimaging data can relay a good impression of the spatial location and extent of an experimental effect, they are also just thresholded images which in many cases do not convey the nature of the effect (e.g. which condition in a interaction effect show the main activation). It is thus necessary to extract and display the real imaging data. SPM offers a variety of different plots for 1st level analyses (e.g. contrast estimates, fitted responses, event-related responses etc.), but unfortunately these are not readily available for 2nd level analyses.

Nowadays however, the main interest in the results from neuroimaging studies lies at the 2nd level. rfxplot is a toolbox intended to easily display the data at the second level (across subjects) in a variety of different plots. It is both suitable for an inspection of the data in the course of different statistical analyses, as well as producing high quality figures for publications.

The toolbox offers a graphical user interface (GUI) for configuring all different plot options (described in detail in the User Manual). Because configuring a plot via the GUI may take some time, the toolbox provides the possibility to reuse the previous plot configuration (for plotting the same effects at a different voxel) without going through the GUI again. In addition, plot configuration can be saved and reused later in a different MATLAB session. Furthermore, the plot configuration can be scripted in a separate m-file and hence rfxplot can be used non-interactively in a batch system.

rfxplot always extracts the data from the underlying 1st level analyses. This way the toolbox avoids some confusion inherent to some rescaling at the 2nd level analyses. This way, for instance, a true estimate of the percent signal change can be plotted.