Before calling the FEAT GUI, you need to prepare each session's data as a 4D Analyze format image (there are utilities in fsl/bin called avwmerge and avwsplit to convert between multiple 3D images and a single 4D - ie 3D+time - image). If the data requires any scanner-specific corrections (for example, for artefacts such as slice dropouts), this should be applied to the data before running FEAT.
Structural images for use as "highres" images in registration should normally be brain-extracted using BET before being used by FEAT.