Neural Signal Processing and/or Interface Methods, Architectures, Apparatuses, and Devices

摘要

Processing a neural signal sequence occurs in accordance with a neural signal spiking model that includes an exponential component (EC) and a polynomial component (PC). The exponential component is correlated with the presence of signal sequence noise, and the polynomial component is correlated with the presence of detectable signal sequence spikes distinguishable from the noise. A neural interface includes a frequency shaping amplifier (FSA) configured for receiving input signals; an amplifier gain stage and an analog-to-digital conversion (ADC) stage; a Hilbert transformer configured for performing a Hilbert transform upon neural signal data received from the ADC stage; a linear regression engine configured for estimating EC parameters and PC parameters corresponding to Hilbert transformed neural signal data; and a neural spike probability estimator configured for generating a neural spike probability map based upon the EC parameters and the PC parameters.