Development and Effectiveness of a Kinect-based Motor Training System for Children with Cerebral Palsy

Cerebral Palsy (CP), a common cause of physical disability in children, often impairs upper extremity (UE) function, which limits children’s participation in activities of daily living and independence. Recently, constraint-induced therapy (CIT) is one of the effective approaches to improve UE function. However, children often feel demotivated during this therapy. Virtual reality (VR) has been confirmed to improve this disadvantage and widely been used for rehabilitation. However, the effectiveness of VR system is still controversial. VR system combining with effective rehabilitation approach, such as CIT, may be a feasible approach for designing cost-effective motor rehabilitation program for children with CP. Thus, this study was aim to: (1) develop a Kinect-based CIT and to evaluate its feasibility and safety; (2) evaluate the effectiveness of Kinect-based CIT. Kinect-based CIT was developed by the principle of CIT, which emphasize on restraining the less affect UE and intensive, repetitive training of the more affected UE. Moreover, the components of motor learning would be considered to design the training protocol. In this study, we conducted two experiments. One was to valid the feasibility of Kinect-based CIT and the other was to confirm the effectiveness of this program for children with CP. In feasibility study, 10 children with hemiplegic CP (HCP) were enrolled. Children played the game for 20 minutes and range of motion (ROM) of shoulder, elbow, wrist and trunk were evaluated. After the gameplay, children completed a questionnaire of their gameplay experience. In addition, we examined the game recorded variables to evaluate the gameplay performance of 16-session-intervention. In effectiveness study, 10 children with CP, aged from 5-12 years old, were recruited in this study. Children were randomly assigned to Kinect-based CIT group or home-based CIT group. All the participants received the same treatment duration: 2-2.5 hours/ day, 2-3 days/ week, for 8 weeks. To evaluate the treatment effects, all children underwent kinematic analysis of unimanual and bimanual reaching tasks and functional outcome measures for activities of daily living before and immediately after the intervention. In feasibility study of 20-minutes gameplay, all of the children increased the amount of use of their more affected limb and reported an enjoyable and safe experience during gameplay. Moreover, in feasibility study of 36-hours gameplay, the output data of two games showed that successful rate and average completed time would improve over time. In preliminary effectiveness study, children with CP in both groups had improvements in UE motor functions. Comparing to Home-based CIT group, Kinect-based CIT group had more improvements in movement time of their more affected UE and better motor planning. Home-based CIT group improved more in peak velocity and bilateral coordination. In conclusions, our developed Kinect-based CIT was a feasible motor rehabilitation program for children with CP. Kinect-based CIT group had greater improvements in motor performance of their more affected UE. Home-based CIT group improved more in bilateral coordination. In brief, these two kinds of therapy were both effective to improve motor performance in children with CP. However, the Kinect-based CIT could also enhance motivation of children and decrease load of therapists. Thus, this program was more feasible to apply in clinic or home.