Increasing the Anatomical Accuracy of Computational Models of Pallidal Deep Brain Stimulation in Parkinson’s Disease.
Researcher: Khidhr Kotaria Presentation Title: Increasing the Anatomical Accuracy of Computational Models of Pallidal Deep Brain Stimulation in Parkinson’s Disease. Research Focus: Biomedical Engineering / Neuroscience School: University of Minnesota Twin Cities Presentation Type: Poster Presentation Abstract: Parkinson’s Disease (PD) is pathophysiologically characterised by the loss of dopaminergic cells in the substantia nigra (D. James Surmeier), and the emergence of both motor and non-motor symptoms. This heterogeneity can be greater understood using computational modelling. Computational models of DBS estimate the pathways activated by the stimulating contacts and can provide insight into relationships between activated neural pathways and their relationships to improvement or worsening of motor signs. Computational models of both Subthalamic Nucleus (STN) and Globus Pallidus (GP) DBS have been created (Linn-Evans), but lack the physiological and anatomical accuracy needed to confidently determine how DBS is impacting motor function. Current deficiencies in anatomical accuracy include straight-line neuron projections, which aren’t a true depiction of neuronal morphology in the brain, they tend to have varying curvatures that straight lines don’t depict. To accommodate this deficiency, a singular axon model with literature-derived branching pattern was generated using MATLAB to depict the morphology of neurons descending from the external GP (GPe) to the internal GP (GPi), which are the main targets for DBS in Parkinson’s. The code is divided into 3 segments, first establishing the singular neuron descending from the GPe to GPi, which has varying nodes at different distances to mimic the probability of branching at specific points on the neuron. Utilising an iterative algorithm that adds a branch to the line before the branching pattern is established. Finally, a theta (Ө) oscillation algorithm is employed, which enables random angle generation (between -π/4 and π/4) at the nodes of each branched line that account for the curvature of the neurons within the model. The single-neuron model is the first step into a full-scale projection of realistic axonal trajectories in parts of the Basal Ganglia, such as the Putamen, Globus Pallidus, and Subthalamic Nucleus that are innervated during Parkinsonian pathophysiology.