Cortical control of a prosthetic device

A method of generating one or more tuning parameters for use in an algorithm that generates positional information used to generate movement commands for controlling a robotic prosthetic device based on firing rates of a plurality of units of a subject, each unit being associated with one or more neurons of the subject, said one or more tuning parameters describing how said firing rates are modulated with velocity, the method comprising:

• for each of a number of trials, collecting:
• (i) an average firing rate of each of said units during each of a number of movement segments of said prosthetic device, and (ii) a plurality of target movement vectors during said number of movement segments of said prosthetic device, wherein each of the target movement vectors represents a normalized displacement of the prosthetic device from an initial state to a target state during the associated movement segment;
• and
• generating the one or more tuning parameters based on the average firing rates collected in each of the number of trials and the target movement vectors collected in each of the number of trials.

A method of controlling a robotic prosthetic device having a gripping element moveable between an opened condition and a closed condition, comprising:

• measuring a set of one or more activity parameters of each of a plurality of units of a subject, each said unit being associated with one or more neurons of said subject;
• employing a real-time extraction algorithm to generate a four-dimensional velocity vector based on each measured set of one or more activity parameters and one or more tuning parameters, said one or more tuning parameters describing how said set of one or more activity parameters are modulated with velocity, said four-dimensional velocity vector having an x direction component, a y direction component, a z direction component, and a gripping element velocity component;
• integrating an appropriately scaled version of each of said x direction component, said y direction component, an said z direction component to obtain an endpoint position and an appropriately scaled version of said gripping element velocity component to obtain a gripping element state;
• and
• using said endpoint position and said gripping element state to generate commands for controlling movement of said prosthetic device and opening and closing of said gripping element.

A method of using a system including a robotic prosthetic device, said system employing an algorithm that generates positional information used to generate movement commands for controlling the prosthetic device based on activity parameters of a plurality of units of the subject, each unit being associated with one or more neurons of the subject, the method comprising:

• calibrating said robotic prosthetic device by generating one or more tuning parameters for use in said algorithm, said one or more tuning parameters describing how said activity parameters are modulated with velocity, said calibrating comprising allowing a user to move said prosthetic device to a plurality of target locations in a physical interaction environment;
• training said user to use said prosthetic device by allowing said user to control movement of said prosthetic device in said physical interaction environment using said algorithm;
• and
• following said calibrating and training, allowing said user to freely control movement of said prosthetic device in said physical interaction environment using said algorithm.

A method of enabling a subject to control a robotic prosthetic device having a plurality of functional portions, each functional portion being associated with a corresponding degree of freedom of movement of said robotic prosthetic device, comprising:

• measuring a set of one or more activity parameters of each of a plurality of units of a subject, each said unit being associated with one or more neurons of said subject;
• predicting said subject's intended movement of said prosthetic device by employing a real-time extraction algorithm to generate a normalized movement vector based on at least each measured set of one or more activity parameters, said normalized movement vector having a plurality of dimensions, each dimension having one or more movement components and being associated with the degree of freedom of a corresponding one of said functional portions;
• for each of said dimensions of said normalized movement vector, scaling the one or more movement components thereof using a scaling factor to convert each of the one or more movement components of the dimension into a corresponding scaled movement component having units appropriate for describing the movement of the one of said functional portions with which the dimension is associated, wherein the units of at least one of the dimensions are different than the units of at least another of the dimensions;
• and
• generating commands for controlling movement of said prosthetic device based on each scaled movement component of each of the dimensions.