Improved detection of event-related functional MRI signals using probability functions

Selecting an optimal event distribution for experimental use in event-relat ed fMRI studies can require the generation of large numbers of event sequen ces with characteristics hard to control. The use of known probability dist ributions offers the possibility to control event timing and constrain the search space for finding optimal event sequences. We investigated different probability distributions in terms of response estimation (estimation effi ciency), detectability (detection power, parameter estimation efficiency, s ensitivity to true positives), and false-positive activation., Numerous sim ulated event sequences were generated selecting interevent intervals (IEI) from the uniform, uniform permuted, Latin square, exponential, binomial, Po isson, chi (2), geometric, and bimodal probability distributions and fixed IEI. Event sequences from the bimodal distribution, like block designs, had the best, performance for detection and the poorest for estimation, while high estimation and detectability occurred for the long-decay exponential d istribution. The uniform distribution also yielded high estimation efficien cy, but probability functions with a long tail toward higher IEI, such as t he geometric and the chi (2) distributions, had superior detectability. The distributions with the best detection performance also had a relatively hi gh incidence of false positives, in contrast, to the ordered distributions (Latin square and uniform permuted). The predictions of improved sensitivit ies for distributions with long tails were confirmed with empirical data. M oreover, the Latin square design yielded detection of activated voxels simi lar to the chi (2) distribution. These results indicate that high detection and suitable behavioral designs have compatibility for application of func tional MRI methods to experiments requiring complex designs. (C) 2001 Acade mic Press.