Functional Engagement in Assisted Therapy Through Exercise Robotics

1. Purpose:

The similarity of functional impairments in the two target hemiplegic populations, teenagers with hemiparetic cerebral palsy CP or acquired brain injury (post-acute phase) and hemiparetic stroke survivors, leads us to pursue a unified rehabilitation strategy. Although age-dependency is clearly a factor in rate of motor learning, intensive therapies have shown similar returns of function with similar intensities and length of treatment across different populations [1]. By involving both adults and children in the same study, the investigators can directly compare the effects of identical interventions using the innovative hardware and software solutions that are developed.

The team of investigators have developed a system that includes a camera (PS3 Eye), 2 modified PS3 Move controllers, and a computer [2]. A software is designed to search through the camera feed to find the glowing balls attached to the two PS3 Move controllers, and changes the position of the computer's mouse cursor based on the position of the two glowing balls. This allows the participants to play simple 2D Facebook games that involve point-and-click by holding on to the Move controllers and moving their arms in space in order to control the cursor and play games.

1. Hypothesis:

The investigators hypothesize that the following strategies will deliver effective, long-term therapy: (1) use a bimanual approach, based on recent research that shows the powerful integrative value of involving both limbs and both cerebral hemispheres in recovery of function post-stroke [3]; (2) target hand-function, following the proven efficacy of this approach (CIMT and GRASP) over exercising arm motions only; (3) directly target patient motivation by embedding the therapy in a social construct of multi-user online gaming such as Facebook Games [4, 5].

1. Justification:

The quality of life of Canadians with stroke declined from 1998 to 2005, a disturbing finding that is directly related to impaired arm function [6]. Traditional (standard) physical therapy includes a range of interventions aimed at improving posture, joint range of motion, muscle tone, muscle strength and motor control. However, no single approach has been proven to have superior effectiveness over any other in upper or lower limb recovery. These are sobering data that illustrate a critical need for novel interventions that will improve function and lead to better quality of life for persons living with hemiplegia.

Motor rehabilitation programs based on low-cost exercise devices such as the Nintendo Wii, Nike Fit and Microsoft Kinect have proven efficacy [5], but participants are required to be in the same room at the same time. Online games such as Farmville on Facebook support cooperative tasks, but there are no appropriate devices to support bimanual input. Our approach will fill this gap by supporting collaborative, bilateral motor-learning that involves force and movement therapy and social media effectively.

1. Objectives:

Our objective is to demonstrate that motivation in adherence to a therapy regimen can be enhanced by the use of online games in conjunction with motion capture devices that require the user to interact bimanually with the game. Our project has 4 phases, each with its own milestones:

1. Start-up: In-clinic focus group meetings with rehabilitation professionals and follow-on focus groups with therapy clients; design of hardware and software based on the outcomes of these focus group meetings (months M1-M9). This phase is finished.
2. Pilot study: Hardware and software development (M10-12), followed by in-clinic usability studies with adult stroke survivors and children with hemiplegia; development of final hardware and software based on the outcomes of the usability studies; integrative Knowledge Transition activities (months M7-M24). This phase is finished.
3. Home-based studies: Two parallel 6-month studies, one with adults and one with teenagers (months M25-M30). The investigators have finished planning of this phase and have received the ethics approval to start recruiting.
4. Knowledge translation: Data analysis, reporting and dissemination (months M30-M36)
5. Research Method (Only Phase 3 is included): The study will have a randomized cross-over design with an early and a late intervention group. A maximum of 20 stroke survivors will be divided into a late intervention and an early intervention group (10 in each group, randomly assigned), and a maximum of 20 teens with hemiplegia will be divided into a late intervention and an early intervention group (10 in each group, randomly assigned). The procedures for late intervention groups are the same. Similarly, the procedures for the early intervention groups are the same. Note that this study is asking the same research question for two different clinical populations.

Each intervention group will be involved in the study for 6 months. Both groups will start with a baseline assessment of outcome measures in week 1. The early intervention group will receive the FEATHERS intervention in weeks 1-8. There will be another round of outcome measure assessments at the end of week 8 (assessing motor recovery). There will be a period of no intervention between weeks 9-16, followed by another round of outcome measure assessments (assessing retention). Finally, weeks 17-24 will also be a no intervention period, followed by another round of outcome measure assessments (assessing variation of the outcomes even without a therapy intervention). The late intervention group will receive 2 months of no intervention followed by a round of outcome measure assessments (assessing variation of the outcomes even without a therapy intervention). Then they will receive the FEATHERS intervention in weeks 9-16. There will be another round of outcome measure assessments at the end of week 16 (assessing motor recovery). Weeks 17-24 will be another period of no intervention followed by another round of outcome measure assessments (assessing retention).

Primary Outcome (functional outcomes) and secondary outcomes (participation outcomes) of this study are as follows:

i.Primary Outcome: Wolf Motor Function Test WMFT (this is the test for which the investigators conducted the a priori power calculation; it is expected to have ~80% chance of detecting an effect with given this sample size).

ii.Secondary Outcomes:

1. Body Structure/Function: Motricity index; grip/pinch strength; NIH stroke scale; modified rankin test; reach performance scale.
2. Activity/Participation: Canadian occupational performance measure; confidence scale; paediatric motivation scale, usability of the system.

More details on the measures are provided under "Outcome Measures".

1. Statistical Analysis:

Phase 3: Based on previous research, the investigators estimate a standardized effect size of 0.60 to 0.80 improvement (Cohen's d) from baseline to terminal assessment on the WMFT. By recruiting 20 stroke subjects and 20 CP subjects, it is expected to have a statistical power (1-β error probability) of approximately 80% for the within-subject change. Our principal statistical test is thus the within-subject t-test comparing pre-intervention WMFT scores to terminal WMFT test scores, combining data from both the early and the late intervention group. In addition to this a priori comparison, WMFT test scores will be analyzed in a Group (early versus delayed intervention) by Time (Weeks 0-8, 9-16, and 17-24) mixed-factorial ANOVA. Secondary outcomes will include motion log files, scoring from the games themselves, functional outcome measures, discussions with participants and exit surveys. Secondary outcomes of function/structure and activity/participation will be analyzed using the same Group by Time mixed-factorial ANOVA. Exploratory comparisons can also be undertaken. For instance, the investigators data analysis will allow us to compare the motor learning between younger and older persons with hemiplegia and the effect of the social media framework on motivation using both quantitative and qualitative metrics (between-subjects t-tests on the functional outcomes and motivation surveys). The investigators will be investigating whether there is a statistical correlation between subjective measures of motivation and the observed changes in functional outcomes.