In this study we tested a rehabilitation paradigm that simultaneously exercised the arm and hand, including the fingers, in an integrated manner using virtual reality task-based gaming simulations. Our goal was to improve hemiparetic hand function in patients in the chronic phase post-stroke. As a group, the subjects were able to more effectively control the upper limb during reaching and hand interaction with the target as demonstrated by improved proximal stability, smoothness and efficiency of the movement path. The improvements in smoothness are indicative of a decrease in the number of sub-movements required to complete the transport phase of the motion. Several authors cite this pattern of change as consistent with improvements in neuromotor control [9, 23]. This improved control was in concert with improvement in the distal kinematic measures of fractionation and improved speed. Of note, these changes in robotic measures were accompanied by robust changes in the clinical outcome measures.
Several factors may have influenced the findings in this study. Congruent with the motor learning and neuroplasticity literature, it is believed that the acquisition of a motor skill follows a dose-response relationship . In rehabilitation, the dose is often measured as the number of task repetitions or practice hours. Multiple authors cite the ability of robotically facilitated training to provide highly repetitious training as a key factor for its effectiveness [30, 31]. The comparison between the training volume typical to robotic interventions and those of traditional UE interventions is marked. Subjects average over 500 repetitions/day in studies in the robotic rehabilitation literature [32–34] while an observational study of the repetitions performed in a traditional outpatient setting averaged 85 . The average number of repetitions during the two to three hour training sessions used in this study exceeded 2200.
Based upon a review of 20 RCT's, it has been suggested that a minimal dose of at least sixteen hours of practice is required to achieve functional changes . Our subjects performed 22 hours of training, 10 hours during week one and 12 hours in the longer sessions in week two. Each training session in this study was considerably longer than the twenty to ninety minute sessions described in the current robotic literature [30, 31] and was delivered within a more concentrated time period [11, 34, 36–38].
Another factor to consider is that the gaming simulations structured the subjects' attentional focus. It has been shown in people with and without disabilities that the learning of a motor task is more effective when attention is focused on externally rather than on internally based directions [39, 40]. In these virtual reality simulations, practice was directed to achieve action goals rather than performing specific movements. The instructions for the game, the feedback provided and the inherent structure of each simulation directed the players' attention to the task to be achieved. In other words, the focus of attention was on the effect of one's movements rather than on the movement itself.
The largest improvements demonstrated with the Virtual Piano were for finger fractionation, which is the ability to flex one finger independently of the other fingers. During practice, the performance feedback, the sound of the appropriate note, occurs when a fractionation target is achieved, reinforcing this construct. In addition fractionation is also specifically reinforced with an adaptive algorithm that increases and decreases the fractionation target, based on the subjects' performance. This algorithm which is described in detail elsewhere appears to help progress the subject towards improved finger function . Subjects made larger improvements in fractionation than speed or accuracy that were not shaped with an algorithm or reinforced with feedback. Similarly, subjects also failed to make improvements in peak finger extension, which was not reinforced with an algorithm, during Hammer Task training. These results are congruent with those of Lum et al.  who found that subjects with strokes, training using the MIME system, reduced force direction errors when this construct was shaped with an algorithm.
Day three training performance for the three proximal kinematic measures (hand deviation, path length and trajectory smoothness), deviates from the trend of daily incremental improvement during the rest of the trial (See Figure 3). Three subjects, all with chronic strokes had their worst performance on day three for these measures. This may be secondary to higher levels of fatigue associated with the initiation of an intense training protocol in these subjects. A comparable pattern of high levels of fatigue during the early days of a trial has been demonstrated by a group of CIMT subjects with chronic strokes .
Our overarching goal is to integrate development of robotic assisted training devices with the most effective training paradigm for recovery of hand function. It is therefore important to compare the changes in JTHF time in this current study to other studies performed in our lab. In a former study of comparable duration, that trained the hand only, the subjects showed a 10% improvement in the time of the JTHF , while in this current study that trained the arm and hand simultaneously, there was a 24 sec decrease in the time to complete the JTHF achieved by the subjects in this study, which was equal to a 20% change in the time needed to complete all the items on the JTHF. This decrease in time represents 27% of the difference between the initial scores of the stroke subjects, and the aged matched controls. Moreover, it represents 33% of the difference between the initials scores of the impaired and uninvolved hand. Given this robust improvement as well as the difference between initial scores for the impaired arm and the less impaired arm, one can suggest that functional changes may have occurred secondary to this training. Future analyses would be required to relate this robust change in the JTHF with changes in activities of daily living function.
Essential factors such as the dosage and intensity of the practice, the focus of attention on the movement outcome, and the drive provided by specific algorithms are important to achieving functional outcomes. However, these factors have been similar in our past studies. What was different in this study was the complexity of the movements required to interact with the virtual simulations. When we trained the hand alone, the gaming simulations were very simple activities, requiring only control of wrist and finger movement. Whereas in this study the activities required by the gaming simulations were more complex and required simultaneous control of integrated shoulder, elbow, forearm, wrist and finger movements. These factors appear to have had a substantial, positive effect on our goal of improving hemiparetic hand function.
However, an important question to consider is whether it is the complexity of the simulations or the consistent training of integrated shoulder, elbow, forearm, wrist and finger movements that is responsible for these improvements. This question engenders another possible training variation. Will the findings be as robust if the subjects train on complex activities that only require independent and discrete upper arm movements or hand movements. To answer this question our lab is in the process of initiating a randomized controlled trial testing for the effect of integrated versus isolated training of proximal and distal upper extremity effectors to compare the outcomes with our previous findings.