Examining Cortical and Subcortical Processes Underlying Speech Auditory-motor Adaptation and Syllable Sequence Learning

Abstract

Sensorimotor control and learning rely on the dynamic interplay of cortical-subcortical circuits, particularly the cerebellum and basal ganglia. Beyond their roles in feedforward and feedback control, these cortical-subcortical circuits are implicated in distinct motor learning mechanisms, including sensory prediction error-based learning, reinforcement learning, strategy-based learning, and motor sequence learning. While most knowledge about cortico-subcortical contributions to motor control comes from studies of non-speech movements, similar mechanisms are thought to underlie speech sensorimotor control and learning. Disorders affecting these structures, such as Parkinson's disease (PD), essential tremor (ET), and cerebellar degeneration, are associated with motor speech impairments, altered speech sensorimotor control, and deficits in speech motor learning. This dissertation integrates behavioral, neurostimulation, and neurophysiological evidence to investigate the cortical and subcortical processes underlying two fundamental forms of speech motor learning: auditory–motor adaptation and syllable sequence learning. Chapter I reviews the literature on speech sensorimotor control in basal ganglia and cerebellar disorders, focusing on motor and sensory impairments, performance in auditory perturbation tasks, and the modulatory effects of pharmacological and deep brain stimulation (DBS) interventions. The review highlights both converging and divergent findings across patient populations and underscores key knowledge gaps. Chapter II presents a behavioral study testing the effects of basal ganglia DBS in PD and cerebellar-related DBS in ET on auditory-motor adaptation and syllable sequence learning. Results showed that these learning processes are differentially affected in PD and ET, but they are not acutely modulated by DBS. ET patients exhibited reduced early adaptation under gradual perturbations and lower overall sequence accuracy, whereas PD patients performed comparably to controls in both domains. Chapter III investigates subthalamic nucleus (STN) electrophysiology during speech production in PD patients using subcortical recordings from implanted DBS electrodes. STN activity showed beta event-related desynchronization/re-synchronization (ERD/ERS) during speech, attenuated under stimulation and modulated by sequence complexity, but unrelated to auditory-motor adaptation, suggesting a role for the STN in speech sequencing. Finally, Chapter IV employs electroencephalography in healthy adults to compare oscillatory dynamics of auditory-motor adaptation and syllable sequence learning. Both tasks elicited alpha/beta ERD during speech planning, but only sequence learning showed stronger alpha ERD in early versus late trials, identifying alpha desynchronization as a potential neural marker of syllable sequence learning. Together, these studies delineate complementary roles of cerebello-thalamo-cortical and basal ganglia-thalamo-cortical circuits in speech sensorimotor learning, clarify how neurological disorders and DBS shape these processes, and identify oscillatory markers that distinguish distinct forms of speech sensorimotor learning.