Optimal experimental design for event-related fMRI

An important challenge in the design and analysis of event-related or singl e-trial functional magnetic resonance imaging (fMRI) experiments is to opti mize statistical efficiency, i.e., the accuracy with which the event-relate d hemodynamic response to different stimuli can be estimated for a given am ount of imaging time. Several studies have suggested that using a fixed int er-stimulus-interval (ISI) of at least 15 sec results in optimal statistica l efficiency or power and that using shorter ISIs results in a severe loss of power. In contrast, recent studies have demonstrated the feasibility of using ISIs as short as 500 ms while still maintaining considerable efficien cy or power. Here, we attempt to resolve this apparent contradiction by a q uantitative analysis of the relative efficiency afforded by different event -related experimental designs. This analysis shows that statistical efficie ncy falls off dramatically as the ISI gets sufficiently short, if the ISI i s kept fixed for all trials. However, if the ISI is properly jittered or ra ndomized from trial to trial, the efficiency improves monotonically with de creasing mean ISI. importantly, the efficiency afforded by such variable IS I designs can be more than 10 times greater than that which can be achieved by fixed ISI designs. These results further demonstrate the feasibility of using identical experimental designs with fMRI and electro-/magnetoencepha lography (EEG/MEG) without sacrificing statistical power or efficiency of e ither technique, thereby facilitating comparison and integration across ima ging modalities. (C) 1999 Wiley-Liss, Inc.