Exploring the neural networks of temporal expectations in rhythm with fMRI: SMA processes pattern-based expectations

Temporal expectations (e.g., predicting “when”) are a key element of musical rhythms, which makes these rhythms an excellent way of looking at temporal expectations in the brain. Temporal expectations can be based on regular, periodic beats, or on predictable patterns with varying interval lengths (e.g., the morse code of the rhythm). However, it is unknown whether beat-based and pattern-based expectations utilize different or shared neural networks. In an fMRI study, 45 participants listened to rhythms and rated their complexity. We used computational models of periodicity (quantifying beat-based expectations) and predictability (quantifying pattern-based expectations) to select rhythms that maximally differed on these aspects. Behaviorally, the absence compared to the presence of beats and patterns both independently increased the perceived complexity of rhythms. Whole-brain and ROI analyses revealed that the cerebellum and the SMA showed increased activity for unpredictable...Show moreTemporal expectations (e.g., predicting “when”) are a key element of musical rhythms, which makes these rhythms an excellent way of looking at temporal expectations in the brain. Temporal expectations can be based on regular, periodic beats, or on predictable patterns with varying interval lengths (e.g., the morse code of the rhythm). However, it is unknown whether beat-based and pattern-based expectations utilize different or shared neural networks. In an fMRI study, 45 participants listened to rhythms and rated their complexity. We used computational models of periodicity (quantifying beat-based expectations) and predictability (quantifying pattern-based expectations) to select rhythms that maximally differed on these aspects. Behaviorally, the absence compared to the presence of beats and patterns both independently increased the perceived complexity of rhythms. Whole-brain and ROI analyses revealed that the cerebellum and the SMA showed increased activity for unpredictable compared to predictable rhythms, which is not in line with previous research. Additionally, contrary to previous studies, there was no change in activity in the medial temporal lobes and basal ganglia based on periodicity and predictability differences. Furthermore, multivariate analyses revealed different activity patterns in the left caudate based on periodicity. However, contrary to previous research, this was not true for the right caudate, SMA, and putamen. Additionally, in line with previous absolute time interval research, the cerebellum and MTL both showed different activity patterns based on predictability differences. However, this was also true for the caudate, putamen, and SMA. These results especially question the role of the SMA in processing temporal expectations, suggesting that beat-based and pattern-based expectations may have been confounded in previous research with factors, such as complexity task demands, and on-beat tones.Show less

Leiden University Student Repository

Documents

In Collections

Exploring the neural networks of temporal expectations in rhythm with fMRI: SMA processes pattern-based expectations