How beat perception coopts motor neurophysiology

Abstract

A fundamental aspect of music cognition is the perception of an underlying metronome-like pulse ("the beat") in complex rhythmic patterns. A key finding from neuroimaging is that even in the absence of movement, beat perception strongly engages the motor system. Existing efforts to create conceptual and computational models of beat perception do not strongly engage with known neurocomputational properties of the motor system. Here we construct a mechanistic model of beat perception grounded in neurophysiology. The model uses two distinct timing signals in medial frontal cortex to infer tempo and to produce intervals that anticipate beats. Signals are sent along the direct pathway through basal ganglia to initiate the covert production of intervals. In the absence of direct pathway activation, medial frontal cortex is driving the hyperdirect pathway, creating a periodic cycle of global inhibition and facilitation of movement initiation. Striatal dopamine reinforces a top-down beat percept when anticipation is accurate and a bottom-up percept when it is not. We highlight the experimental results and lines of reasoning that support these hypotheses; we implement the model in silico and note important aspects of human beat-based timing that it reproduces; and we explore the model's predictive and explanatory power, especially regarding the therapeutic effect of the beat in Parkinson's disease. Highlights Box• The Action Simulation for Auditory Prediction (ASAP) hypothesis posits that covert anticipatory timing of a musical beat is performed by the motor system. • Current discussion of beat-based timing does not address the balance between external responsiveness and internal maintenance necessary to adapt to beat timing changes but maintain a beat through syncopated or ambiguous rhythms. • Beats may be anticipated by an adjustable-speed ("relative") timer in the supplementary motor area, initiated and reset like a movement via the basal ganglia direct pathway. • The beta oscillation that fluctuates with the beat may be an anti-kinetic "hold" signal that would scaffold entrainment of non-specific movements.• Striatal dopamine may serve as an index of confidence in the beat tempo and phase, allowing the motor system to transition between a synchronization and beat-continuation.Abbreviations BBT = beat-based timing MFC = medial frontal cortex PMC = premotor cortex IBS = "internal beat strength" SCT = synchronization/continuation task 108 (49), 19784-19789. https://doi.org/10.1073/pnas.1112933108 6 Note that their absolute/relative terminology does not align with ours: we [unfortunately] categorize their "relative timing cells" and "absolute timing cells" both as "relative timing signals."

Inverted central auditory hierarchies for encoding local intervals and global temporal patterns

“…It is interesting that cortical spiking timescales are sensitive to a stimulus feature that is approximately 40 3 slower than the timescale itself (800 ms versus 20 ms), which suggests that A1 regular-spiking unit timescale dynamics are likely the product of other local or long-range circuits that respond more directly to the change in predictable stimulus timing. 27 Prior work has shown that the temporal decay of sensory-evoked spikes can be regulated by recurrent local circuits and inter-regional feedback loops. 6,7,[28][29][30][31] Cortical GABAergic neurons that express somatostatin or neuron-derived neurotrophic factor are clear candidates to dampen principal neuron spiking timing because they target the distal dendrites of A1 pyramidal neurons to regulate network excitability and recurrent excitation, but are also sensitive to slowly changing internal state variables.…”

Expectancy-based rhythmic entrainment as continuous Bayesian inference

“…Though PIPPET and PATIPPET are not committed to a particular brain-based implementation, advances in the brain basis of timing and beat-keeping combined with the hypothesized neural bases of predictive processing suggest the beginnings of a plausible implementation of PIPPET in the brain. A detailed discussion of a possible neural basis of beat maintenance is presented in [63]. Briefly, supplementary motor area may maintain an ongoing estimate of mean phase through some combination of intrinsic dynamics and interaction with the basal ganglia, while dopaminergic signaling in striatum may maintain an estimate of phase uncertainty.…”

“…Experiments with non-human primates have shown neural trajectories in medial premotor cortex (MPC, encompassing the supplementary and pre-supplementary motor areas) that represent progress through self-generated behavioral processes. The author hypothesizes in [76] that similar trajectories represent rhythmic phase in human MPC. A representation of a linear phase ϕ , used in the phase inference framework for flexibility and mathematical tractability, would seem to be a limiting factor for implementation in the brain.…”

“…Guided by the “Action Simulation for Auditory Prediction” (ASAP) hypothesis presented in [80] and further developed in [76], the theory of hierarchical predictive processing [81], and the predictive functions proposed for the dorsal auditory pathway [82, 83], we propose a neural implementation of PIPPET’s phase estimation in Figure 8. An essential aspect of this account is that it does not insist on the mathematical convenience of instantaneous phase updates, which are obviously implausible in the brain.…”

Mapping the contents of consciousness during musical imagery

“…Whereas it is commonly hypothesized that motor commands lead to better prediction of sensory inputs and thus benefit individual's adaptability to its environment (Wolpert et al, 1995;Schubotz, 2007;Schroeder et al, 2010;Cannon and Patel, 2020), the advantage of such a mechanism is not as clear during imagery, in which the sensory information had ostensibly already been established and can be retrieved at will. Nevertheless, the undeniable gap between stored knowledge and its transient representation in conscious experience could perhaps be somewhat mitigated by this mechanism.…”