• beta bursts
• cycle-by-cycle
• Parkinson's
• speech
• subthalamic nucleus

This thesis investigates the detection of oscillatory bursts in local field potentials through a time-domain algorithm and examines their modulation during a speech production task in Parkinsonian patients. Beta frequency transient activity, called bursts, is a well-established biomarker for Parkinson’s disease, correlating with symptoms and dopamine dynamics. Speech impairments are common in this disorder; however, the role of STN-cortico connections in speech production are not yet fully understood. I analyzed intracranial cortico-subthalamic simultaneous data from patients undergoing deep brain stimulation (DBS) implantation surgery, compared the performances of three burst detection approaches, and investigated burst features evolution before, during and after speech. Employing a cycle-by-cycle burst analysis allowed for a more reliable output and obtaining more informative features, notably burst frequency and time symmetries.
My study contributes to a comprehensive beta activity characterization, encompassing speech-related modulations, with the intent to advance towards a more effective and personalized therapy for Parkinson’s disease and a refinement in scientific knowledge about a complex, uniquely human behaviour that is speech.