Development of Computational Models of Pedunculopontine Nucleus Stimulation for Clinical Trials and Mechanistic Studies

Abstract

Deep brain stimulation (DBS) in the pedunculopontine nucleus (PPN), a component of the mesencephalic locomotor region in the brainstem, has been proposed to alleviate gait and balance disturbances associated with Parkinson’s disease; however, clinical trials results have been highly inconsistent. Such variability may stem from inaccurate targeting in the PPN region, modulation of fiber pathways implicated in side effects, and lack of understanding of the modulatory effects of DBS in the brainstem. Here, we describe the development and refinement of computational models that can predict the neuromodulatory effects of PPN-DBS in both the non-human primate and human. These models included (1) brain atlas-based models that combined detailed biophysically realistic neuron and axon models with a finite element model simulating the voltage distribution in the brain during DBS, (2) high-field 7T MRI techniques to visualize and create volumetric morphologies of structures in the brainstem for use in the models, and (3) clinically relevant subject-specific computational models that incorporate the anisotropic conductivity of the brain tissue. Based on the validated results of these models, we can conclude that the neuronal pathways modulated by DBS in the brainstem are highly sensitive to both lead location and stimulation parameters. These computational models of DBS will be useful in future clinical trials, both prospectively to plan DBS lead trajectories and improve stimulation titration and retrospectively to investigate the underlying mechanisms of therapy and side effects of stimulation.