Validation of a time-scaling-based model for representation of dynamics in humans and its applications in rehabilitation

The generation of muscle forces to perform various human movement tasks is a complex control problem whose mathematical solution requires complex formulations due to redundancy in muscle actuators. This problem becomes increasingly difficult because the same movement can be generated by following different joint coordination strategies. Literature suggests that the human brain resolves these redundancies using an optimization-based approach. Such approaches, in general, require systems with large computation capabilities to generate a feasible solution. The human brain, however, has seemingly developed a method to solve this problem in a simpler manner. This study proposes a simpler time-scaling-based mathematical model of the dynamics of human movement that eliminates the need for the optimization routines. It is proposed that humans separate the muscle forces required to compensate for gravitational effects and to generate motion. In the framework of the proposed model, the muscle forces required to generate motion for one given speed are scaled to generate the same motion at different speeds. This theory of movement control is applicable under the following three assumptions: (1) Humans have the ability to separate muscle forces required for compensation of gravitational effects from total muscle forces. (2) Time scaling is a characteristic of human movement. (3) The scaling-based solution satisfies the conditions of optimality.;Various aspects of this theory of movement control are tested by applying force fields with a custom robotic device. The adaptation of the subjects to the force fields is used to validate the theory. Experiments conducted with 54 healthy individuals provide evidence suggesting that humans may be separating muscle forces required to compensate for gravity from the total muscle forces required to perform motion. The experiments also suggest that scaling is a primary characteristic of human movement that is manifested as various kinematic invariants for human movement reported in the literature. Methods are developed with the same device for assessing rehabilitation during the acute phase of recovery from stroke. Various metrics are developed with the device, and their applicability was tested by comparing arm movement patterns of 8 older and 8 younger healthy subjects. Preliminary results with 3 stroke patients show that the methods developed can be used to assess post-stroke recovery. Application of the model proposed for classifying stroke impairments is proposed, and its applicability in classification and rehabilitation of lost motor function remains to be tested in a clinical setting with experiments involving stroke patients.