The fMRI happened 3 hours after the participants dose, so sleep vs. no sleep wouldn't affect the results (you would reasonably expect the sleep vs. no sleep to be randomly distributed between participants, without being unduly biased to one cohort).
Study participants received a single-dose of modafinil or a placebo pill.
All subjects then underwent two fMRI scans, performed before and 3 h after
drug (or placebo) administration.