Brain-machine interface utilizing interventions to emphasize aspects of neural variance and decode speed and angle using a kinematics feedback filter that applies a covariance matrix
A brain machine interface (BMI) for restoring performance of poorly performing decoders is provided. The BMI has a decoder for decoding neural signals for controlling the brain machine interface. The decoder separates in part neural signals associated with a direction of movement and neural signals associated with a speed of movement of the brain machine interface. The decoder assigns relatively greater weight to the neural signals associated with a direction of movement.
1. An artificial controller of a prosthetic device, comprising:
(a) a brain machine interface having an algorithm executable by a computer, wherein the brain machine interface comprises a mapping from neural signals to corresponding intention estimating kinematics of a limb trajectory, wherein the intention estimating kinematics comprises speeds and angles;
b) the brain machine interface controlling the prosthetic device using recorded neural signals as input to the brain machine interface and the mapping of the brain machine interface determining the intention estimating kinematics to control the prosthetic device, wherein the controller results in an executed movement of the prosthetic device;
(c) a modified brain machine interface having an algorithm executable by the computer for modifying during the control of the prosthetic device, at discrete time intervals over the course of the executed movement of the prosthetic device, the angles in the brain machine interface, wherein each of the modifications of the angles comprises changes in the direction of the angles towards an end target of the executed movement of the prosthetic device;
(d) the modified brain machine interface controlling the prosthetic device; and
(e) a kinematics feedback filter executable by the computer in both the brain machine interface and the modified brain machine interface, wherein the kinematics feedback filter applies a covariance matrix of an a posteriori estimate of the intention estimating kinematics at each time step of the discrete time intervals, whereby the kinematics are modeled as feedback from the interfaces to a user of the prosthetic device.
2. The artificial controller as set forth in claim 1 , further comprising an algorithm for projecting the neural signals into a space of lower dimensionality compared to the dimensionality of the neural signals, and re-weighting the contribution of the dimensions of the lower dimensional space to the estimated kinematics.