Frequency shifts and depth dependence of premotor beta band activity during perceptual decision-making

Chandramouli Chandrasekaran, Iliana E. Bray, Krishna V. Shenoy

Abstract

Neural activity in the premotor and motor cortex shows prominent structure in the beta frequency range (13-30 Hz). Currently, the behavioral relevance of beta band activity (BBA) in premotor and motor regions is not well understood. The underlying source of motor BBA and how it changes as a function of cortical depth is also unknown. Here, we addressed these unresolved questions by investigating BBA recorded using laminar electrodes in the dorsal premotor cortex (PMd) of two male rhesus macaques performing a visual reaction time (RT) reach discrimination task. We observed robust BBA before and after the onset of the visual stimulus but not during the arm movement. While post-stimulus BBA was positively correlated with RT throughout the beta frequency range, pre-stimulus correlation varied by frequency. Low beta frequencies (~15 to 20 Hz) were positively correlated with RT and high beta frequencies (~25 to 30 Hz) were negatively correlated with RT. Simulations suggested that these frequency-dependent correlations could be due to a shift in the component frequencies of the pre-stimulus BBA as a function of RT, such that faster RTs are accompanied by greater power in high beta frequencies. We also observed a laminar dependence of BBA, with deeper electrodes demonstrating stronger power in low beta frequencies both pre- and post-stimulus. The heterogeneous nature of BBA and the changing relationship between BBA and RT in different task epochs may be a sign of the differential network dynamics involved in expectation, decision-making, and motor preparation.

Introduction

Beta band activity (BBA), typically defined as oscillations between 13 and 30 Hz, is a hallmark of the motor system. Traditionally, BBA has been associated with the "maintenance of the status quo," where its presence indicates a state of postural stability and its suppression precedes movement. However, this monolithic view fails to account for the complex dynamics of beta oscillations during decision-making and motor preparation. This study investigates whether BBA is composed of distinct sub-bands with unique behavioral correlates and how these signals are distributed across the laminar architecture of the dorsal premotor cortex (PMd).

Methods

Two rhesus macaques (Monkey T and Monkey O) were trained on a visual discrimination task. They viewed a checkerboard of red and green squares and performed a reaching movement toward a target matching the dominant color. The difficulty of the task was modulated by the color coherence of the checkerboard, which introduced variability in reaction times (RT). Behavioral performance was modeled using a psychometric function to relate the probability of a correct choice to the stimulus coherence.

Psychometric function used to model choice performance as a function of color coherence.

Neural recordings were obtained using 16-contact laminar electrodes (U-probes) inserted into the PMd, allowing for simultaneous sampling across the cortical depth (~2.4 mm). This setup enabled the differentiation between superficial (input-related) and deep (output-related) layers. The researchers analyzed the correlation between beta power and RT using partial Spearman correlations, controlling for stimulus difficulty. Additionally, synthetic Local Field Potential (LFP) data were simulated to test the "frequency shift" hypothesis.

Results

Frequency-Dependent Behavioral Correlates

The study revealed that BBA is not a uniform signal. During the pre-stimulus epoch, the relationship between beta power and RT was frequency-dependent: low beta (~15–20 Hz) was positively correlated with RT (higher power associated with slower responses), whereas high beta (~25–30 Hz) was negatively correlated with RT (higher power associated with faster responses). In contrast, during the post-stimulus epoch, the entire beta range showed a positive correlation with RT.

The Frequency Shift Hypothesis

Simulations supported a "frequency shift" model, suggesting that as the motor system prepares for a faster response, the center frequency of pre-stimulus beta oscillations shifts toward higher frequencies. This shift explains the opposing correlation patterns observed in the low and high beta sub-bands.

Laminar Distribution of Beta Activity

Analysis of the cortical depth revealed a clear laminar dependence. Power in the low beta range (10–20 Hz) was significantly stronger in the deeper electrodes (corresponding to Layer V) compared to superficial layers. Given that Layer V contains the primary output neurons of the motor cortex, this suggests that low-frequency beta oscillations are closely linked to motor output or feedback mechanisms.

Discussion

These findings suggest that premotor beta activity is a heterogeneous signal with dynamics that vary by frequency, task epoch, and cortical depth. The shift in pre-stimulus frequency suggests that the motor system actively tunes its oscillatory state to facilitate rapid action. The concentration of low beta in deep layers points toward a specific role in the output circuitry of the PMd. While the deep-layer dominance of low beta was robust across subjects, high beta results showed more variability, potentially reflecting anatomical differences. Future research integrating spiking activity with these laminar LFP patterns will be essential to fully understand the circuit-level mechanisms of motor decision-making.

Summary updated

Abstract

ABSTRACT Neural activity in the premotor and motor cortex shows prominent structure in the beta frequency range (13-30 Hz). Currently, the behavioral relevance of beta band activity (BBA) in premotor and motor regions is not well understood. The underlying source of motor BBA and how it changes as a function of cortical depth is also unknown. Here, we addressed these unresolved questions by investigating BBA recorded using laminar electrodes in the dorsal premotor cortex (PMd) of two male rhesus macaques performing a visual reaction time (RT) reach discrimination task. We observed robust BBA before and after the onset of the visual stimulus but not during the arm movement. While post-stimulus BBA was positively correlated with RT throughout the beta frequency range, pre-stimulus correlation varied by frequency. Low beta frequencies (~15 to 20 Hz) were positively correlated with RT and high beta frequencies (~25 to 30 Hz) were negatively correlated with RT. Simulations suggested that these frequency-dependent correlations could be due to a shift in the component frequencies of the pre-stimulus BBA as a function of RT, such that faster RTs are accompanied by greater power in high beta frequencies. We also observed a laminar dependence of BBA, with deeper electrodes demonstrating stronger power in low beta frequencies both pre- and post-stimulus. The heterogeneous nature of BBA and the changing relationship between BBA and RT in different task epochs may be a sign of the differential network dynamics involved in expectation, decision-making, and motor preparation.
BibTeX
@article{chandrasekaran2019frequency,
  author = {Chandramouli Chandrasekaran and Iliana E. Bray and Krishna V. Shenoy},
  title = {{Frequency shifts and depth dependence of premotor beta band activity during perceptual decision-making}},
  journal = {Journal of Neuroscience},
  year = {2019},
  doi = {10.1523/jneurosci.1066-18.2018},
}
Figure