How to test for dual-task-specific effects in brain imaging studies--an evaluation of potential analysis methods

The study of the concurrent performance of two tasks allows deep insights into the human cognitive system and, accordingly, an increasing number of brain imaging studies are conducted to identify the neuroanatomical correlates of such dual-task performance. In this overview we present currently used approaches to identify dual-task-specific activations in fMRI and PET studies. A comparison is made in order to identify the approaches which have the potential to validly detect dual-task-specific activation patterns, i.e. activation which cannot be explained by the individual performance of the component tasks alone. We demonstrate that while all approaches suffer from at least some drawbacks, the best (although potentially over-conservative) approach is to compare the dual task with the sum of the single tasks, the second-best is an interaction contrast, and the third-best a conjunction analysis. Comparisons of the dual task with the mean of single-task activity or with only one single task should be avoided except for a few specific situations. We generalize our conclusions to related research areas, such as multisensory integration or divided attention.