||<100% bgcolor="#FFFF99">'''Using a field-map to undistort EPIs'''<<BR>>[[FmBackground|Background]]<<BR>>[[FmPower|Improvement in power!]]<<BR>>[[FmFieldmaps|Acquiring fieldmaps]]<<BR>>[[FmDownload|Download]]<<BR>>[[FmRequirements|What you'll need to undistort an EPI data series]]<<BR>>[[FmReference|Guide and reference]]<<BR>>[[FmExample|Step-by-step example]]<<BR>>[[FmBatch|Batch mode]]||

= Improvement in power in RFX analysis =
What you really care about is probably power to detect functional activation. Here are the results of a random effects analysis (N=12) of a motor cortex localiser task. Subjects were asked to press a button either with a finger of their left hand (left column) or their right (right column). The data were analysed with or without undistortion.

{{http://imaging.mrc-cbu.cam.ac.uk/images/motortask.gif}}

Functional images (thresholded at 0.001 uncorrected) are shown superimposed on the SPM T1 template. The numbers in the top left of each image indicate peak t values over the whole volume. After the application of undistortion, clear activation can be seen in motor cortex (BA 4) on each side.