Abstract
The growing number of recording sites of silicon based probes means that an increasing amount of neural cell activities can be recorded simultaneously, facilitating the investigation of underlying complex neural dynamics. In order to overcome the challenges generated by the increasing number of channels, highly automated signal processing tools are needed. In this paper we present ELVISort, a deep learning method which combines the detection and clustering of different action potentials in an end-to-end fashion. We show that the performance of ELVISort is comparable with other spike sorting methods which use manual or semi-manual techniques, while exceeding the methods which use an automatic approach. We also demonstrate that ELVISort can exploit the massively parallel processing capabilities of GPUs via deep learning frameworks, with the potential to be used on other cutting-edge AI specific hardware such as TPUs.