Abstract
This study describes novel methods for navigating and placing of electrodes into specific structures in the basal ganglia for deep brain stimulation (DBS), as it is common in the treatment of Parkinson's disease. Critical to these procedures in neurosurgery is the localization and identification of different target structures such as subthalamic nucleus (STN) along the electrode's trajectory and finding the best position for the stimulating electrode. Typically, microelectrode recordings (MER) of local neural activity along up to five parallel trajectories are used by neurosurgeons for detecting the target region and creating the anatomic positions of the electrodes by imagination. We developed a method for automatic classification of the MER signals, which provides an electrode model with patient specific borders of the STN. In addition, a method is provided for finding the best matching of the electrode model with a 3D model of the STN. As a result, a 2.5D visualization of the target region is produced with the most probable positions of the electrodes and their intersections.