| A mechatronic structural model of the mammalian muscle spindle Ia response was developed and used to investigate neuromechanical mechanisms contributing to individual spindle dynamics and the information content of spindle ensemble response. Engineering specifications were derived from displacement, receptor potential and Ia data in the muscle spindle literature, allowing reproduction of core muscle spindle behavior directly in hardware. A linear actuator controlled by a software muscle model replicated intrafusal contractile behavior; a cantilever-based transducer reproduced sensory membrane depolarization; a voltage-controlled oscillator encoded strain into a frequency signal. Results of engineering tests met all performance specifications. Data from the biological literature was used first to tune the model against 5 measures of ramp and hold response, then to validate the fully tuned model against ramp and hold, sinusoidal and fusimotor response experiments. The response with dynamic or static fusimotor input was excellent across all studies. The passive spindle response matched well in 5 of 9 measures. Dynamic intramuscular strain data from 28 locations on the surface of a contracting rat medial gastrocnemius was sent sequentially through the model to reconstruct the Ia ensemble response of a large population of muscle spindles. Results showed that under dynamic fusimotor stimulation, the ensemble significantly increased Ia correlation to whole muscle kinematic inputs and that homogeneously distributed dynamic fusimotor stimulation increased Ia ensemble correlation to muscle velocity in a dose-dependent manner. Proposed mechanisms include decorrelation of spindle noise by intramuscular strain inhomogeneities and fusimotor-dependent noise and nonlinear gains, as well as fusimotor-dependent velocity selectivity. Potential applications for the robotic model include basic science motor control research and applied research in prosthetics and robotics. |
