Biologically realistic simulated extracellular potential data as benchmark for the evaluation of spike detection algorithms

Many spike detection and spike sorting algorithms are found in the literature. Their performance is either compared on the basis of experimental data or on the basis of synthetic data, mostly generated by the random distribution of spike templates in time and the addition of noise. But never before biologically realistic simulated extracellular potential data was used as benchmark, providing both, control and realistic features. We will describe the generation of our simulated test data within a biologically realistic network simulation and finally present some performance results of selected spike detection methods applied to our data.