Abstract
Electroencephalography (EEG) allows for source measurement of electricalbrain activity. Particularly for inverse localization, the electrode positionson the scalp need to be known. Often, systems such as optical digitizingscanners are used for accurate localization with a stylus. However, theapproach is time-consuming as each electrode needs to be scanned manually andthe scanning systems are expensive. We propose using an RGBD camera to directlytrack electrodes in the images using deep learning methods. Studying andevaluating deep learning methods requires large amounts of labeled data. Toovercome the time-consuming data annotation, we generate a large number ofground-truth labels using a robotic setup. We demonstrate that deeplearning-based electrode detection is feasible with a mean absolute error of5.69 +- 6.1mm and that our annotation scheme provides a useful environment forstudying deep learning methods for electrode detection.