Robotically facilitated virtual rehabilitation of arm transport integrated with finger movement versus isolated training of the arm and hand in persons with hemiparesis

Introduction. Stroke is the leading cause of disability in the United States affecting 200 to 300 per 100,000 inhabitants. Eighty to 95 percent of these persons demonstrate residual upper extremity impairments lasting beyond six months after their strokes. Robotically facilitated therapeutic activities, performed in simple and complex virtual environments have emerged as one approach to improving UE rehabilitation outcomes after stroke. The goals of this dissertation were: (1) To compare the effects of robotically facilitated integrated upper extremity task practice to that of a similarly applied, robotically facilitated program of training the upper extremity effectors in isolation on clinical measurements of upper extremity motor function, in persons with upper extremity hemiparesis secondary to stroke, (2) To compare the effects of robotically facilitated integrated upper extremity training to those of a similarly applied program of isolated upper extremity training on motor learning and motor control, utilizing robotically collected kinematic and performance measurement of training activities in persons with upper extremity hemiparesis secondary to stroke, (3) To compare the effects of robotically facilitated integrated upper extremity training to those of a similarly applied program of isolated upper extremity training on home-based upper extremity activity in persons with upper extremity hemiparesis secondary to stroke. Methods. A double-blind randomized controlled trial with repeated measures performed prior to training and three days after training. Forty individuals post-stroke, participated in a two week training protocol using the NJIT-RAVR system, a six-degrees of freedom haptic robotic system and the NJIT TrackGlove systems a data system designed to track three-dimensional reaching and finger movement behavior. These two systems allowed the participants to interact with one of two programs of complex virtually simulated rehabilitation activities. The first program, hand and arm separate (HAsep) training, employed simulations in which research participants performed activities controlled by finger movement only, using the NJIT-Track Glove system and activities controlled by arm movement only utilizing the NJIT-RAVR system. Hand and am together (HAtog) training incorporated simulations controlled by a combination of arm and finger movement. Primary outcome measures included three clinical tests of upper extremity movement, the Wolf Motor Function Test (WMFT), the Jebsen Test of Hand Function (JTHF) and the Nine Hole Peg Test (NHPT). Secondary measures included seven kinematic measurements that were collected by the NJIT-RAVR and NJIT-Track Glove systems during training. Exploratory outcome measurements were three metrics used to evaluate out of the laboratory, upper extremity activity, measured prior to and after training using ActiGraph triaxial accelerometers. Results. Hypothesis testing identified negligible differences between adaptations to HAsep and HAtog training as measured by the clinical tests of upper extremity function utilized as primary measures in this study. These findings were consistent when considering the entire battery of tests grouped as a whole, as well as the individual test scores produced by the WMFT, JTHF and NHPT. Each of the two training programs elicited improvements in kinematic measurements collected during training. HAsep training was more effective for improving measurements related to the proximal musculature's role as a primary mover of the entire upper extremity. HAtog training was more effective for improving measurements of finger individuation ability and the proximal musculature's role as a stabilizer of the upper extremity in space during distal musculature's interaction with objects. A small sample size and technical challenges limited the conclusions that could be drawn from twenty four hour activity measurements and the changes in these measures subsequent to training as measured in this study. This said, a possible relationship between changes in the total amount of roll plane movement performed in a day by persons with hemiparesis secondary to stroke and changes in hand dexterity subsequent to training as measured by the Jebsen Test of Hand Function was identified. Significance. Both of the training protocols evaluated in this study produced robust changes in clinical tests of UE motor function, providing clear support for the ability of robotically facilitated training in virtual environments to produce changes in untrained motor function as measured using clinical tests of motor function. These changes exceeded published MCID and MDC for the tests utilized and exceeded the outcomes reported for several major trials of technology based as well as non-technology based, upper extremity RTP in persons with stroke. (Abstract shortened by UMI.).