RESUMO
Paced Auditory Serial-Addition Task (PASAT) is a complex task commonly used to examine patients with diffuse brain damage. A visual version of the neuropsychological test (Paced Visual Serial-Addition Task, PVSAT) has also been introduced to clinical practice, and both versions were adapted to be used in neuroimaging, namely functional magnetic resonance imaging (fMRI). The aim of our work was direct comparison of auditory and visual versions of the paced serial addition test (PASAT/PVSAT) in a within-subject and within-session study and description of the commonalities and differences in both activated and deactivated brain regions. Twenty young adult right-handed healthy volunteers participated in the study and underwent whole-brain fMRI examination during PASAT and PVSAT performance. Higher-level statistical analysis was performed to generate group mean activation and deactivation maps for both tasks, their conjunctions and differences across modalities. In PASAT/PVSAT activation conjunction analysis, we confirmed the existence of a modality-independent neural network similar to working memory tasks and to previous PASAT or PVSAT studies. In PASAT/PVSAT deactivation conjunction analysis, we observed a rather symmetrical extensive pattern of deactivated regions, overlapping the default mode network. Significant differences between PASAT and PVSAT were found in the right frontal eye field (FEF) and bilaterally in the striate and extrastriate cortices. Activation in one task and deactivation in the other jointly contributed to significant differences in all occipital and occipitotemporal regions. Both tasks activated right FEF, but activation during PASAT was significantly stronger than during PVSAT. Between-modality differences should be considered when preparing and interpreting neuroimaging experiments.