Biomimetic and Non-biomimetic Extraction of Motor Control Signals Through Matched Filtering of Neural Population Dynamics

2 Brain-machine interfaces rely on extracting motor control signals from brain activity in real time 3 to actuate external devices such as robotic limbs. Whereas biomimetic approaches to neural 4 decoding use motor imagery/observation signals, non-biomimetic approaches assign an arbi5 trary transformation that maps neural activity to motor control signals. In this work, we present 6 a unified framework for the design of both biomimetic and non-biomimetic decoders based on 7 kernel-based system identification. This framework seamlessly incorporates the neural popula8 tion dynamics in the decoder design, is particularly robust even with short training data records, 9 and results in decoders with small filter delays. The theory and results presented here provide a 10 new formulation of optimal linear decoding, a formal method for designing non-biomimetic deco11 ders, and a set of proposed metrics for assessing decoding performance from an online control 12 perspective. The theoretical framework is also applicable to the design of closed-loop neural 13 control schemes. 14