Bioelectronic Medicine and the Dawn of Robotic Training to Improve Motor Outcome in Chronic Stroke
- 17 Downloads
Engineers and clinicians have cooperated to produce and test new classes of bioelectronics that have altered motor impairment that occurs after stroke. The rationale that increased intensity of training alters outcome derives from past clinical and preclinical work. Now several studies have demonstrated that interactive robotic devices are a potent tool for the therapist to deliver effortlessly, reproducible high intensity movement training. These robots are safe and can provide a platform so that recovery might be influenced by a combination of noninvasive novel treatment programs. Also, these robotic devices provide a continuous objective history of movement parameters that will open the horizon for their use in generating novel movement biomarkers to understand, predict and measure the influence of new treatments on motor outcome after neurological injury.
The enormous personal and societal burden caused by diseases of the brain and spinal cord make imperative innovative attempts to reduce illness and alter permanent disability. Colleagues at MIT, HI Krebs and N Hogan, developed an array of interactive robotic devices that we have used to aid and abet treatment programs for neurological recovery of motor function of the limbs in patients who have had a stroke (1, 2, 3, 4). These interactive robots move a patient’s paretic arm and when the patient begins to move, these robots “get out of the way” so the patient can execute the movement with very little resistance from the device. With these robots, a therapist can generate training sessions that are intensity controlled (a single one-hour session requires over 1,000 to and fro movements of a limb segment). The robots are tireless agents that produce reliable, reproducible movement sequences. In addition, the controllers on these robotic devices can be tuned to individual patients so that the robot can present different physical challenges at the point when a patient is moving the robot arm, so that training can focus, for example, on the speed, trajectory (aiming) or force of a movement (5,6). The success of controlled multicenter randomized studies that used robotic protocols to improve the outcome of upper limb motor recovery in patients with chronic stroke prompted the American Heart Association to make robotic training standard care (5,7, 8, 9). This work will review the continued relationship between rehabilitation robotics for the paretic upper extremity after chronic stroke and focus on frequent questions that arise in clinical practice: Can intensive training alter the dreaded “performance plateau”? Can the human-robot interaction be optimized for an individual patient? Can explicit training of the affected limb generalize to improvement on untrained motor tasks? Can the control group be trained in an intensive manner to mimic the robot training? Can the robot-derived measurements of movement provide an objective biomarker for studies of other treatments, especially new pharmacology for acute stroke? Can the robot training emerge as a platform for combination therapy?
Can Intensive Robotic Training Alter a Performance Plateau?
The movement of the curves rightward toward higher Fugl-Meyer values is significant (Kolmogorov-Smirnoff, p < 0.001). The details are instructive and show that for patients with severe impairment and scores in the 20s, the improvement occurred in the first nine treatments, yet for those with moderate impairment and scores in the 30s, the improvement occurred throughout the training. Moreover, for some with this moderate impairment, there was further improvement after the training ceased. That a performance plateau should be considered reaching an optimum or having diminished capacity for improvement should give way to more refined individualized and varied treatment approaches.
Can the Human-Robot Interaction be Optimized for an Individual Patient?
Improved movement after robot therapy depends on the kinematic details of the robot training.
Controller tuned to:
Change: admission to dischargea
Can Explicit Training of the Affected Limb Generalize to Improvement on Untrained Motor Tasks?
The data show that motor performance may generalize beyond the trained tasks. The clinically significant changes support patient anecdotes of being able to use eating utensils more effectively, being able to bath and dress independently, and, of course, there is always a sportsperson, in our case, an avid golfer, who returned to the links.
Can the Control Group be Trained in an Intensive Manner to Mimic the Robot Training?
Essentially intensive motor training improves outcome for the paretic limb, and whether this activity is delivered by a therapist or a robot appears to make little difference, except to the therapist.
Can the Robot-Derived Measurements of Movement Provide an Accurate Objective Relevant Biomarker?
Much as robotic devices are critical adjuncts to deliver intensive treatment in stroke recovery trials, clinical research protocols in stroke, especially for drug outcome studies, would be aided by a reliable, repeatable and temporally efficient continuous measure of impairment. Recently we were able to measure the movement of the paralyzed upper limb in a longitudinal manner in 208 patients who were recovering from stroke (23). We tested the relevance of the relationship among four standard and well-known clinical movement impairment scales and objective robot measures. For the clinical scales, therapists were trained and interand intrarater reliability was assured (24). For the robot measure we used a series of visually guided and visually evoked unconstrained reaching and attempts to move against resistance (provided by the robot) (25). From these movements, we were able to calculate deviation from a straight line, aiming, average, peak speed and duration of movement, and also characterize the movement smoothness. The robotic measures predicted well the clinical measures, so for the standard Fugl-Meyer impairment assessment scale the R2 = 0.73, and for the standard Motor Power scale the R2 = 0.75. These results suggest that the robot measures of motor performance adequately capture outcome and make the robotic device a useful tool that provides objective, repetitive, reliable and speedily obtained measurements of motor performance.
Can the Robot Training Emerge as a Platform for Combination Therapy?
Real world rehabilitation techniques have been influenced by preclinical experiments and clinical observations that have demonstrated the felicitous effects of “intensive” motor training before the dawn of the “electroceutical” revolution in robotics (27, 28, 29, 30). Similarly, clinical investigators have produced important trial information to show that dedicated stroke units where therapists deliver intensive movement programs provide the best outcome (20,31,32). And, finally, these experimental findings are in accord with the detailed anecdotal reports of particular patients who have documented extraordinary motor recovery after stroke that appeared to depend on effort and intensive practice (33, 34, 35).
Stroke is the number one cause of permanent disability (36). And this dubious leadership position is unlikely to recede because surprisingly high incidence information for stroke is driven by an aging population and better epidemiology from urban populations, and recent equally surprising prevalence information is driven by improved survival after the acute event, and also the emerging use of the only effective treatment for acute stroke, tissue plasminogen activator.
The conclusions regarding the problem of treating persistent permanent neurological impairment has to focus on new technology. Since the answer to each argument presented here is, resoundingly, yes, it behooves rehabilitation enterprises to adopt robotic devices as important adjuncts to the current therapy protocols. Moreover, as the science of clinical recovery proceeds to be measured by waves of small molecules, or novel drugs, or cell-based therapies that repair or alter central nervous system networks, then the ultimate phenotypic motor reorganization will require retraining. With robotics, therapists are finally armed with tools that will deliver reproducible, intensity controlled retraining alone, or in combination with other invasive or non-invasive brain stimulation. And finally, these robotic devices will provide objective measurements for any study of motor recovery.
The authors declare that they have no competing interests as defined by Bioelectronic Medicine, or other interests that might be perceived to influence the results and discussion reported in this paper.
BT Volpe was supported from NIH RO1-HD069776, the Buchanan Family Foundation and the Feinstein Institute for Medical Research.
- 12.Krebs HI. (1997) Robot-aided neuro-rehabilitation and functional imaging [dissertation]. Cambridge (MA): Massachusetts Institute of Technology.Google Scholar
- 14.Palazzolo JJ. (2005) Robotic technology to aid and assess recovery and learning in stroke patients [dissertation]. Cambridge (MA): Massachusetts Institute of Technology.Google Scholar
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, and provide a link to the Creative Commons license. You do not have permission under this license to share adapted material derived from this article or parts of it.
The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
To view a copy of this license, visit (https://doi.org/creativecommons.org/licenses/by-nc-nd/4.0/)