Portable and wearable electromyographic biofeedback for spinal cord injury to enhance neuroplasticity

A method, comprising:

• receiving, by a processor, electromyography data representing electrical signals resulting from muscle activation by a user;
• receiving, by the processor, inertial data representing movement or spatial orientation of a plurality of inertial sensors associated with the user, wherein the inertial data comprises accelerometer data, gyroscope data, and magnetometer data;
• measuring, by the processor, using the accelerometer data, the gyroscope data, and the magnetometer data, a joint rotation angle about a hinge joint axis while compensating for multi-axial human joint rotation by:
• constructing based on quaternion pose signal data from the one or more wearable sensor devices, a rotation vector incorporating a unit norm of the quaternion pose signal data,
• projecting the rotation vector along a first principal component representative of the hinge joint axis of rotation as a unit norm vector, and
• measuring the joint rotation angle about the hinge joint axis based on a length of the rotation vector following projection;
• updating, by the processor, an attitude vector associated with each inertial sensor using the accelerometer data, the gyroscope data, and the magnetometer data, wherein gyroscopic drift is compensated for by using a set of measured accelerometer and magnetometer data and a set of predicted accelerometer and magnetometer data;
• determining, by the processor, a joint angle vector between two inertial sensors of the plurality of inertial sensors using attitude vectors associated with the two inertial sensors, wherein a baseline attitude vector for each of the two inertial sensors is used to establish relativity between the two inertial sensors;
• determining, by the processor, a rotation rate around a joint axis defined by the joint angle vector;
• and
• controlling actions in an experience presented to the user based on timing and intensity of muscle activation of the user and one or more of the electromyography data, the attitude vector for each inertial sensor, the joint angle vector between the two inertial sensors, or the rotation rate around the joint axis.

A tangible, non-transitory, computer-readable medium having instructions encoded thereon, the instructions, when executed by a processor, being operable to:

• access electromyography data representing electrical signals resulting from muscle manipulation including amplitude and timing of a muscle contraction by a user associated with at least one wearable sensor device;
• access inertial data representing movement or spatial orientation associated with the at least one wearable sensor device, wherein the inertial data comprises accelerometer data, gyroscope data, and magnetometer data;
• map a set of qualities of the electromyography data to actions of the user in a gaming experience, wherein the set of qualities of the electromyography data comprise timing and intensity of muscle activation;
• control actions of the user in the gaming experience based on the set of qualities of the electromyography data;
• and
• measure, using the accelerometer data, the gyroscope data, and the magnetometer data, a joint rotation angle about a hinge joint axis while compensating for multi-axial human joint rotation by:
• constructing, based on quaternion pose signal data from the wearable sensor device, a rotation vector incorporating a unit norm of the quaternion pose signal data;
• projecting the rotation vector along a first principal component representative of the hinge joint axis of rotation as a unit norm vector;
• and measuring the joint rotation angle about the hinge joint axis based on a length of the rotation vector following projection.