Using MVPA to disentangle the role of familiarity and predictability in rhythm perception

Poster Session F - Tuesday, April 1, 2025, 8:00 – 10:00 am EDT, Back Bay Ballroom/Republic Ballroom

Jessica Grahn^1,2 (jgrahn@uwo.ca), Joshua Hoddinott¹; ¹Centre for Brain and Mind, Western University, ²Department of Psychology, Western University

Listening to rhythmic sounds elicits activity in the basal ganglia, cerebellum, and premotor and supplementary motor areas. Rhythms with a strong beat elicit more SMA and basal ganglia activity relative to non-beat, irregular rhythms. However, most music that people listen to has a clear beat, thus neural responses to strong-beat rhythms may be driven by familiarity rather than beat perception per se. To address this confound, we trained participants on a subset of rhythms that varied in beat strength, and measured BOLD activity to the rhythms before and after familiarity was increased via this training. Participants trained on 12 unique rhythms (4 strong-beat, 4 weak-beat, and 4 non-beat) in finger tapping tasks. In pre- and post-training fMRI sessions, we measured BOLD responses while participants listened to the rhythms during a rhythm discrimination task. Preliminary results reveal no influence of familiarity on the neural representation of beat – strong-beat and non-beat rhythms elicited significantly dissimilar activity patterns in the SMA and putamen both in pre-training and post-training scans. The only effect of training was an increase in dissimilarity between activity patterns in these regions, suggesting that learning the rhythms only made activity more distinct between beat strength conditions. Overall, this suggests the SMA and putamen do not simply encode the familiarity or predictability of rhythms, but are likely involved in beat perception itself.

Topic Area: PERCEPTION & ACTION: Other