The Newman Lab

An initial therapy protocol for 42 chronic stroke patients (≥ 3 months post-stroke) consisted of robot-assisted moves to and from 8 targets that are oriented like a compass, starting with the north target and proceeding clockwise, i.e., N, NE, E, SE, etc. This therapy resulted in statistically significant reductions in shoulder and elbow impairment and shoulder joint pain [1]. The effects of the therapy were specific to the muscles exercised and did not generalize to the patient’s wrist and hand. In addition, the presence of neural plasticity after brain injury suggests that the physiological processes underlying motor recovery are similar to those underlying motor learning [2-4]. Similar to motor learning, motor recovery should be enhanced by adapting the amount of robot assistance based on patient performance.

Robot-assistance during the initial therapy was provided by a point impedance controller that defines a virtual spring between the patient hand position, y, and a desired virtual trajectory, yvt (Figure 1 with N target at y=L and yvt=0.25L). Although the robot assists patients who are unable to move, it also retards motion of patients who are capable of moving ahead of yvt. An adaptive slot impedance controller was developed to allow capable patients to move ahead of yvt (Figure 2, also at yvt=0.25L). In addition, the time allotted to move from one target to another, tm, and the stiffness normal to the target axis, ksw, were varied based on an algorithm that utilized measures of patient performance [5]. Note, tm is an indicator of a patient’s ability to move along the target axis, whereas ksw is an indicator of the ability to aim along the target axis.

As reported in [6], thirty patients with hemiplegia caused by a stroke that occurred at least 8 months prior to their initial evaluation participated in the PBPT protocol. Figure 3 demonstrates the evolution of the control parameters during early and late therapy sessions for a patient. This patient tended to move slower (tm ↑), but aim better (ksw ↓) as the early session progressed. Notice, the patient was producing quicker, better-aimed movements by the late therapy session. Although more patients need to complete the PBPT protocol to make a stronger statement, the reduction of clinical impairment measures from these patients appears to be many times greater (a factor of four to a factor of ten [6]) than the reductions from the initial protocol. Under the working hypothesis that motor recovery resembles motor learning, the initial success of the PBPT protocol provides further evidence that robotic therapy works by enhancing neural plasticity.

Bibliography

[1] S. E. Fasoli , H. I. Krebs, J. Stein, W. R. Frontera, R. Hughes, and N. Hogan, “Robotic therapy for chronic motor impairments after stroke: follow-up results,” Arch Phys Med Rehabil, vol. 85, no. 7, pp. 1106-1111, 2004.

[2] C. S. Li, C. Padoa-Schioppa, and E. Bizzi, “Neuronal correlates of motor performance and motor learning in the primary motor cortex of monkeys adapting to an external force field,” Neuron, vol. 30, no. 2, pp. 593-607, 2001.

[3] R. J. Nudo, B. M. Wise, F. Sifuentes, and G. W. Milliken, “Neural substrates for the effects of rehabilitative training on motor recovery after ischemic infarct,” Science, vol. 272, no. 5269, pp. 1791-1794, 1996.

[4] R. A. Schmidt and T. D. Lee, Motor Control and Learning: A Behavioral Emphasis, Champaign, IL: Human Kinetics, 1999.

[5] H. I. Krebs, J. J. Palazzolo, L. Dipietro, M. Ferraro, J. Krol, K. Rannekleiv, B. T. Volpe, and N. Hogan, “Rehabilitation robotics: performance-based progressive robot-assisted therapy,” Auton Robot, vol. 15, no. 1, pp. 7-20, 2003.

[6] M. Ferraro, J. J. Palazzolo, J. Krol, H. I. Krebs, N. Hogan, and B. T. Volpe, “Robot-aided sensorimotor arm training improves outcome in patients with chronic stroke,” Neurology, vol. 61, no. 11, pp. 1604-1607, 2003.