Detection power, estimation efficiency, and predictability in event-related fMRI

Experimental designs for event-related functional magnetic resonance imagin g can be characterized by both their detection power, a measure of the abil ity to detect an activation, and their estimation efficiency, a measure of the ability to estimate the shape of the hemodynamic response. Randomized d esigns offer maximum estimation efficiency but poor detection power, while block designs offer good detection power at the cost of minimum estimation efficiency. Periodic single-trial designs are poor by both criteria. We pre sent here a theoretical model of the relation between estimation efficiency and detection power and show that the observed trade-off between efficienc y and power is fundamental. Using the model, we explore the properties of s emirandom designs that offer intermediate trade-offs between efficiency and power. These designs can simultaneously achieve the estimation efficiency of randomized designs and the detection power of block designs at the cost of increasing the length of an experiment by less than a factor of 2. Exper imental designs can also be characterized by their predictability, a measur e of the ability to circumvent confounds such as habituation and anticipati on. We examine the relation between detection power, estimation efficiency, and predictability and show that small increases in predictability can off er significant gains in detection power with only a minor decrease in estim ation efficiency. (C) 2001 Academic Press.