Identification of Reflex and Intrinsic Mechanical Changes Associated with Spasticity

Spasticity (reflex hyperexcitability and hypertonus) and contracture are major sources of disability in neurologically impaired patients. The hypertonus and reflex hyperexcitability disrupt the remaining functional use of muscles, impede motion, and may cause pain. Prolonged spasticity may be accompanied by structural changes of muscle fibers and connective tissue, which may result in a reduction in joint range of motion (ROM) and lead to a clinical contracture. The overall objectives of this study are to characterize both reflex and intrinsic mechanical changes in spastic muscles through in vivo experiments and innovative identification methods, to gain insights into the mechanisms of spasticity, to develop a new rehabilitation modality and evaluate treatment outcome quantitatively.

Specially designed small-amplitude random perturbations are applied to the spastic joint over a range of initial ankle flexion angles, to characterize reflex and intrinsic properties of muscles and other soft tissues crossing the joint. A nonlinear delay differential equation model is used to characterize neuromuscular dynamics in terms of the dynamic stretch reflex gains for stretching and shortening muscles, static stretch reflex gain, joint stiffness, joint viscosity, and foot inertia. By these means, we distinguish intrinsic changes in muscles and connective tissues from changes in reflex pathways.

Fig. 1. The block diagram of a joint neuromuscular control system including both intrinsic mechanical and reflex mechanisms. The system input, DT(t), and output, Dq(t), are deviations in the joint torque and angle, respectively. I(l), B(l) and K(l) are limb inertia, joint viscosity and stiffness, respectively. DT(t) overcomes the limb inertia torque DTi(t), and the remaining torque DTm(t)=DT(t)-DTi(t) is what applied to muscles and sensed by the GTO’s. DTg(t) is the GTO mediated force-feedback torque. The lower portion of Fig. 1 corresponds to the spindle reflex pathway, and DTs(t) is the spindle-mediated reflex torque. DTp(t) is used to overcome the joint viscous and elastic resistance. rp(l) and rn(l) are the dynamic stretch reflex gains for muscle lengthening and shortening, respectively. Kd(l) is the static stretch reflex gain. td1 and td2 are the spindle- and GTO-mediated reflex loops delays, respectively. The dashed branch of force feedback is omitted for the simplified models discussed below.

On the other hand, to minimize intrinsic contributions and manifest reflex contributions to joint torque, tendon reflexes are evaluated with several new measures to investigate mechanisms of hyperactive reflexes in spasticity. A hand-held instrumented hammer or a computer controlled linear actuator is used to tap the involved tendon and record the tapping force, while joint torque and muscle EMG signals are recorded isometrically as measures of the reflex responses. Tendon tapping force (designated as system input), reflex torque (as output), their dynamic relationship (characterized as system parameters tendon reflex gain, contraction rate, and reflex loop delay) are used to evaluate tendon reflexes.

The system parameters provide more repeatable measures than do input or output parameters alone because they quantify the input and output simultaneously and dynamically. Despite significant muscle weakness, spastic patients show significantly higher reflex gain, contraction rate, and shorter reflex loop delay, indicating hyperexcitability of motoneurons and peripheral receptors, and a relatively larger portion of the muscle activated reflexively with greater recruitment of larger fast-twitch fibers. Furthermore, spastic patients show significantly lower threshold in tapping force, yet their evoked reflex torque is significantly higher. Clinically, the system measures also correlate more closely with Ashworth scale and tendon reflex scale than do the output measures.