Direct electrical stimulation of the premotor cortex shuts down awareness of voluntary actions

Understanding the conscious and unconscious processes behind willed actions is a major challenge in neuroscience. We often move our bodies according to our desires, yet the process involves a chain of unconscious events culminating in a conscious experience of action. A key insight into the anatomical basis of motor awareness comes from neuropsychological disorders like anosognosia for hemiplegia (AHP), where patients believe they are moving a paralyzed limb. An established model suggests that lesions to the premotor cortex (PMC) disrupt motor awareness by preventing the detection of a mismatch between intended and unexecuted movements. This suggests PMC's role in a comparator system matching expected and actual motor outputs. This study directly tests the PMC's role in motor execution and awareness using direct electrical stimulation (DES) in awake patients undergoing brain surgery. Twelve patients performed a rhythmic hand-manipulation task (HMt), with eight also performing an online verbal motor-monitoring task (MMt) evaluating real-time performance. Four additional patients used a delayed MMt due to potential speech difficulties during stimulation. The PMC and primary somatosensory cortex (S1) were targeted for DES. If the PMC functions as a comparator, disrupting its activity should impair motor awareness, while S1 stimulation should only affect movement execution. We hypothesized that PMC disruption would lead to erroneous reports during the MMt due to impaired awareness, while S1 disruption would only affect the HMt but not awareness.

Previous research on anosognosia for hemiplegia (AHP) has highlighted the importance of the premotor cortex (PMC) in motor awareness. Lesions in the PMC and surrounding areas have been associated with impaired motor awareness, suggesting that the PMC plays a crucial role in a comparator system that matches intended and actual motor outputs. This comparator system is hypothesized to compare an efference copy of the motor command with sensory feedback from the moving limb. Studies have shown that motor awareness can be independent of somatosensory feedback, suggesting that the PMC is involved in a predictive model of motor action rather than relying on direct sensory information. The existing literature provides strong support for the hypothesis that the PMC is a crucial area for motor awareness, and this study aims to further investigate this relationship through direct electrical stimulation.

Twelve awake patients with left-hemisphere brain tumors undergoing surgery participated. They performed a hand-manipulation task (HMt) involving rhythmic manipulation of a cylindrical handle. Eight patients performed an online motor-monitoring task (MMt) providing real-time verbal feedback ('OK' or 'STOP'). Four patients performed a delayed MMt due to possible speech impairment. Direct electrical stimulation (DES) was applied to the premotor cortex (PMC) and the primary somatosensory cortex (S1) as a control. The DES parameters (intensity and frequency) were adjusted to produce a motor arrest (as determined through electromyography (EMG) and video recording). Patients' verbal responses during the MMt were recorded. Electrophysiological monitoring (EEG, ECOG, EMG) was used to monitor cortical and muscular activity during the tasks and stimulation. Two versions of the MMt were used: an online version requiring real-time feedback and a delayed version with retrospective reporting. The stimulation sites were precisely located using neuronavigation and 3D MRI reconstruction. Statistical analysis included Fisher's exact test and Kruskal–Wallis test to compare the effects of DES on the MMt and EMG activity, respectively. A disconnection analysis was conducted using both virtual and actual lesions to investigate the neural networks involved in motor awareness.

Direct electrical stimulation (DES) of both the premotor cortex (PMC) and the primary somatosensory cortex (S1) effectively interrupted the hand-manipulation task (HMt), as indicated by electromyography (EMG) and video recordings. However, only stimulation of the PMC significantly altered the patients' motor awareness. In 88.9% of PMC trials disrupting the HMt, patients reported online that they were executing the action correctly, despite the complete movement arrest. Conversely, during S1 stimulation, patients accurately reported the motor arrest (100%). This effect was replicated in four additional patients with a delayed MMt (100% reported correct execution despite the arrest). Statistical analyses (Fisher's exact test, p < 0.001) confirmed a significant association between altered motor monitoring and PMC stimulation. EMG analysis showed that DES in both PMC and S1 reduced muscle activity compared to baseline, confirming the motor disruption (Kruskal-Wallis test, p = 0.000 for both). Disconnection analysis revealed that the arcuate fasciculus and the superior longitudinal fasciculi II and III were involved in both anosognosia for hemiplegia patients and in PMC virtual lesions. In contrast, S1 virtual lesions primarily involved U-shaped fibers between S1 and M1, a different network.

The findings support the hypothesis that the premotor cortex (PMC) plays a crucial role in conscious motor monitoring. The disruption of motor awareness during PMC stimulation, despite the presence of motor arrest in both PMC and S1 stimulation, strongly indicates the PMC's specific contribution to conscious awareness of voluntary actions. This selective effect aligns with the clinical observation of anosognosia for hemiplegia (AHP), where patients lack awareness of their motor deficits. The disconnection analysis further corroborates this by identifying shared white matter tracts between AHP patients and the PMC virtual lesions. The study provides compelling evidence for the involvement of the ventral PMC, particularly in the dorsal sector, in motor awareness. While other areas, such as the cerebellum and posterior parietal cortex, contribute to motor monitoring, the current findings highlight the PMC's unique role in conscious awareness of motor execution.

This study demonstrates that direct electrical stimulation of the premotor cortex (PMC) selectively impairs conscious awareness of voluntary actions, even when motor execution is disrupted in other areas like the primary somatosensory cortex (S1). The findings strongly support the PMC's role as a crucial hub in the network for human motor awareness. Future research could explore the precise contribution of different PMC subregions and investigate the interaction between the PMC and other brain regions involved in motor awareness, potentially using larger sample sizes and more complex experimental designs.

The study had a relatively small sample size, and the clinical constraints of awake brain surgery limited the experimental design. The simplified MMt focused only on explicit motor awareness, neglecting implicit aspects. The study included only left-hemisphere patients, limiting generalizability. Future research with larger samples and less constrained experimental parameters will strengthen these findings.