Co-Adaptive Neural Interfaces Vocabulary

Vocabulary flashcards covering key terms, mathematical formulations, and experimental components from the research on co-adaptive neural interfaces.

Neural Interface

A system that restores or augments human capabilities by converting high-bandwidth biological signals into control signals for an external device via a decoder algorithm.

Co-Adaptive System

A system in which both the user and the decoder algorithm adapt simultaneously and dynamically in response to one another during closed-loop interactions.

Decoder Algorithm

A computational algorithm that translates a user's biological input signals (such as surface electromyography) into movement control outputs (such as cursor velocity).

SmoothBatch

An adaptive decoding algorithm that periodically updates the decoder matrix online using a weighted combination of the prior decoder and an optimal decoder computed over recent data.

EMG Tuning Curve

A mathematical characterization of user muscle behavior that estimates the relationship between surface electromyography (EMG) activity and the direction of intended movement.

User Encoder ($E$)

The closed-loop mapping between presented task information (target position, target velocity, position error, and velocity error) and the output surface electromyography (EMG) activity.

Potential Game

A game-theoretic framework in which the optimization dynamics of both interacting agents (user and decoder) can be described and analyzed using a single scalar potential function $\phi(E, D)$.

Stationary Points ($E^*, D^*$)

Joint user-decoder strategies in a co-adaptive system defined by zero cost gradients ($\frac{\partial \phi}{\partial E} = 0$ and $\frac{\partial \phi}{\partial D} = 0$), representing equilibrium points where interface parameters persist over time.

Error Decay Rate

The spectral radius (largest eigenvalue magnitude) of the linearized adaptation dynamics about a stationary point, defining the rate of system convergence over iterations.

Decoder Effort Penalty ($\lambda_D$)

A regularization weighting factor in the decoder cost function that penalizes decoder matrix magnitude ($\|D\|^2$), influencing both decoder transformation gain and user muscle effort.

Wiener Filter Decoder

A velocity-controlled linear mapping decoder that transforms processed multi-channel electromyography signals ($u_t$) into cursor velocity output ($v_t = D \cdot u_t$).

Quattrocento System

The multi-channel electrophysiological recording system used to acquire high-density surface electromyography (EMG) signals from the forearm.