Amplitude Adaptive Modulation of Neural Oscillations Over Long-Term Dynamic Conditions: A Computational Study.

Closed-loop deep brain stimulation (DBS) shows great potential for precise neuromodulation of various neurological disorders, particularly Parkinson's disease (PD). However, substantial challenges remain in clinical translation due to the complex programming procedure of closed-loop DBS parameters. In this study, we proposed an online optimized amplitude adaptive strategy based on the particle swarm optimization (PSO) and proportional-integral-differential (PID) controller for modulation of the beta oscillation in a PD mean field model over long-term dynamic conditions. The strategy aimed to calculate the stimulation amplitude adapting to the fluctuations caused by circadian rhythm, medication rhythm, and stochasticity in the basal ganglia-thalamus-cortical circuit. The PID gains were optimized online using PSO, based on modulation accuracy, mean stimulation amplitude, and stimulation variation. The results showed that the proposed strategy optimized the stimulation amplitude and achieved beta power modulation under the influence of circadian rhythm, medication rhythm, and stochasticity of beta oscillations. This work offers a novel approach for precise neuromodulation with the potential for clinical translation.

Functional dynamics of thalamic local field potentials correlate with modulation of neuropathic pain.

Understanding the functional dynamics of neural oscillations in the sensory thalamus is essential for elucidating the perception and modulation of neuropathic pain. Local field potentials were recorded from the sensory thalamus of twelve neuropathic pain patients. Single and combinational neural states were defined by the activity state of a single or paired oscillations. Relationships between the duration or occurrence rate of neural state and pre-operative pain level or pain relief induced by deep brain stimulation were evaluated. Results showed that the occurrence rate of the single neural state of low-beta oscillation was significantly correlated with pain relief. The duration and occurrence rate of combinational neural states of the paired low-beta with delta, theta, alpha, high-beta or low-gamma oscillations were more significantly correlated with pain relief than the single neural states. Moreover, these significant combinational neural states formed a local oscillatory network with low-beta oscillation as a key node. The results also showed correlations between measures of combinational neural states and subjective pain level as well. The duration of combinational neural states of paired alpha with delta or theta oscillations and the occurrence rate of neural states of the paired delta with low-beta or low-gamma oscillations were significantly correlated with pre-operative pain level. In conclusion, this study revealed that the integration of oscillations and the functional dynamics of neural states were differentially involved in modulation and perception of neuropathic pain. The functional dynamics could be biomarkers for developing neural state-dependent deep brain stimulation for neuropathic pain.