MARS Research Project R1:
Evaluating Error Augmentation for Neurorehabilitation
Executive Summary of Aims:
One promising form of robotic training that leverages the power of neuro-plasticity is error augmentation . In this paradigm the computer singles out and magnifies the subject’s movement errors from a desired trajectory, thus forcing the subjects to strengthen their control. This feedback is sometimes counterintuitive and differs greatly from the standard level of care. However, error augmentation requires knowledge of the desired (or intended) movement, which is difficult to presume during complicated functional 3-dimensional (3D) activities. This sub-project evaluates two practical approaches for solving this problem of determining and exploiting the desired movement for rehabilitation. Recent and unique developments in our lab using a state-of-the-art display system interfaced with a haptic robot make it possible to both know the desired trajectory and amplify movement errors in real-time with dramatic results. In recent work from the previous funding cycle we have already shown significant improvements in stroke survivors’ recoveries 1-5. In this proposal we plan to expand this area of research in several brief interventional studies. Our previous RERC funding permitted us to develop a unique, 3D large-workspace system, which uses a haptics (robotic forces) and graphics (visual display) interface (VRROOM). This promising line of work will be extended to the following two clinical aims:
Aim 1: Therapist - driven trajectories. We intend to determine clinical efficacy of several types of therapist-assisted error augmentation on retraining the nervous system in functional activities. One significant challenge is the difficultly in determining the desired trajectory the patient should adopt when practicing complex functional activities. One promising solution is an adaptation of a current rehabilitation training where a therapist specifies the trajectory in real time. This technique also allows the expert therapist to customize their approach to therapy, focusing on what is critical for a particular patient’s recovery. For example, if a subject is having difficulty moving in a certain part of the workspace, the therapist can direct all practiced movements to that region. While there are several lines of compelling evidence that error augmentation is beneficial for stroke, the parameters for how to optimally use error augmentation are not fully known. We will test three experimental treatments in a crossover design: each subject will receive, in randomized order, a control treatment with no error augmentation; a visual error augmentation, or combined visual and haptic error augmentation. We hypothesize that combined haptic and visual error augmentation will lead to the best functional recovery.