Real time hardware neural spike amplitude extraction

A novel algorithm for population spike (PS) amplitude extraction suitable for real time hardware processing was developed. The extraction method was implemented digitally and experimentally tested on a field programmable gate array (FPGA) device using 16-bit quantization. The accuracy of the implementation was tested using PS signals recorded from hippocampal slices. The PS response of the dentate gyrus granule cells were generated in a multi-electrode array (MEA) setup. Spike amplitudes extracted in real time by the hardware were compared with values resulting from floating-point computation in software. Results showed successful implementation of hardware algorithm with average normalized mean square error (NMSE) less than 2%.