Broad Alpha-rhythm Enhancement and Its Convergence during Sequential Memory Task

Alpha-rhythm enhancement during memory maintenance has been explained as active inhibition of task-irrelevant visual inputs because the enhancement is generally observed in occipital visual areas. Moreover, in a sequential memory task, items encoded at an early stage must be maintained in parallel to the encoding of subsequent memory items. To reveal the role of the alpha-rhythm in multiple memory processes, magnetoencephalograms of 14 young participants (age, mean +/- SD = 21.1 +/- 1.1 years) were recorded during a sequential memory task. Participants memorized seven sequentially presented arrow directions. Arrow directions were randomized in the memory trials, whereas they were fixed in the control trials. The time course was divided into four periods: beginning (0-1.2 s: 1st and 2nd arrows presented), midterm (1.2-3.0 s: 3rd-5th arrows), ending (3.0-4.2 s: 6th and 7th arrows), and maintenance (4.2-5.2 s: before recall cue presentation). The source amplitudes of the alpha-rhythm were analyzed by three-way repeated-measures ANOVA (Memory/ Control x Brain region [n = 68] x Time period [n = 4]) with post hoc analysis. In many brain regions, alpha-rhythm amplitude was significantly enhanced in the Memory condition than in the Control condition. These enhancements were distributed widely across brain regions in the beginning period but gradually converged to occipital areas toward the maintenance period. Our results suggest that even if alpha-rhythm enhancement is caused by active inhibition, it is not always solely attributable to the inhibition of visual inputs. Additionally, we compared the time courses of alpha-rhythm amplitude obtained in young participants in this study with those recorded in older participants in a previous study. In the previous work, it was suggested that precuneus activity was essential for memory performance in older people. However, for the young participants in this study, precuneus activity was highly inhibited and not related to memory performance.