Conf Proc IEEE Eng Med Biol Soc 2019 Jul;2019:4221-4224
The subthalamic nucleus (STN) is a commonly used target in deep brain stimulation (DBS) to control the motor symptoms of Parkinson's Disease (PD). Identification of the spiking patterns in the STN is important in order to understand the neuropathophysiology of PD and can also assist in electrophysiological mapping of the structure. This study aims to provide a tool for grouping these firing patterns based on several extracted features from the spiking data. Single neuronal activity from the STN of PD subjects was detected and sorted to compute the binary spike trains. Several features including loca variation, bursting index and the prominence of the peak frequency of the power spectrum were extracted. Clustering of spike train segments was performed based on combination of features in 3D space to scrutinize how well they describe different firing regimes. The results show that this approach could be used to automate the grouping of stereotypic firing patterns in STN.