Neural network dynamics in Parkinson's disease

Parkinson's disease (PD) is characterized by the cell death of neuronal brain cells producing the signaling molecule dopamine. Due to resulting shortage of dopamine, the dynamics of neuronal cells changes, most notably abnormal synchronization of neuronal activity. Such changes complicate the information processing in the brain, resulting in symptoms such as tremor, rigidity and slowness of movement. Deep brain stimulation (DBS) is a surgical treatment where an electrode is implanted to stimulate a specific brain region. DBS is a well-established treatment when medication is no longer effective for PD. DBS is meant to desynchronize pathological oscillations, as they are thought to be the main cause of the symptoms. Despite the high clinical success rate, the way how the pathological activity originates in the brain and how DBS can compensate it are still unresolved questions. Computational modeling is a valuable tool for finding answers to these questions. In the first part of the thesis, computational models are employed in order to get insight in new proposed stimulation therapies for PD. It is demonstrated that stimulation of the pedunculopontine nucleus can eliminate the pathological activity from the entire network model. It is suggested that short-duration desynchronizing stimulation protocols may also disrupt pathological synchronous activity. The results of simulation show that plasticity within the globus pallidus pars externa might be an explanation for this claim. The second half of this thesis focuses on the analysis of single-unit recordings of subthalamic nucleus (STN) cells obtained from PD patients and the acquisition of local field potentials (LFP) in parkinsonian rats. Although it was possible to record clean LFP data, using these data in combination with spiking neuron models is not straightforward. It has been shown that the firing behavior of single units is different in the sensorimotor part of the STN than in other parts of the STN. Postoperative evaluation of target stimulation areas in the investigated PD patients with DBS shows a significant preference for the sensorimotor part of the STN. Therefore, analysis of the firing behavior may help to discriminate the STN sensorimotor part for the optimal placement of the DBS electrode.