Flex2Play: An Ankle Strength Training Device with Virtual Reality Interface for Children with Cerebral Palsy (Washington University in St. Louis)


Increasing ankle strength can improve gait and motor control in children with cerebral palsy, thereby improving quality of life (1).  However, current therapies are clinic-based and it can be difficult to motivate children to comply with prescribed strengthening exercises.  The Flex2Play is an in-home ankle strength training device designed to meet this need.  The Flex2Play has a virtual reality interface that converts patients’ ankle movements into control of online video games through the use of a Nintendo WiiMote.  The device has been designed, built, and demonstrated to be an engaging therapy that allows progressive resistance training and can be customized by a therapist for each user’s needs.


Cerebral palsy (CP) is a non-progressive neurological disorder characterized by impaired motor function and caused by a defect or lesion during brain development (2).  Diagnosed in 3 to 4 children out of every 1,000 live births, CP is the most common developmental disorder associated with lifelong motor impairment and disability (3).  The associated motor impairments include reductions in gait kinematics, muscle strength, motor function and quality of life compared to healthy controls (4).

Due to the high prevalence of CP and the fact that there is currently no cure, improving motor function of children with CP is a high priority.  Recent studies indicate that increasing ankle strength may normalize gait patterns and improve motor control (1).  Clinic-based muscle strengthening programs have achieved positive outcomes but are expensive and the associated travel time may burden patients and interfere with usual social activities (5).  In-home strength training would resolve these concerns but it presents additional challenges.  In particular, the number of repetitions required for effective ankle strengthening is large such that children may bore and fail to adhere to the prescribed training regimen.  Additionally, there is currently no widely-available commercial device that addresses these concerns.


The goal of this project was to meet the observed need by designing an ankle strength training device for children with CP that:
•    Is safe for in-home use
•    Requires minimal setup or training
•    Is affordable (Target price $200)
•    Is portable (Maximum weight 13.6 kg)
•    Motivates patient participation

In addition to these primary goals, the device was designed to fulfill several additional metrics:
•    Allow a full range of motion (-50° to 30° from neutral)
•    Provide adequate resistance (Maximum 120 N dorsiflexion, 360 N plantarflexion)
•    Provide progressive resistance (Minimum 15% increments)
•    Accommodate various patient heights (Foot 20 to 30 cm off ground)
•    Accommodate various foot sizes (Widths 8 to 13.5 cm, lengths up to 30 cm)


To meet the observed need, we designed the Flex2Play.  The Flex2Play is a device that motivates participation in therapy by transforming the ankle movements into control of online video games.   The Flex2Play employs elastic resistance bands to provide a range of resistances against ankle dorsiflexion (foot rotates to move toes toward shin) and plantarflexion (foot rotates to move toes away from shin) and to facilitate progressive resistance training that may be individualized by a therapist for each patient’s needs.  A Nintendo WiiMote strapped to the device connects via Bluetooth to a personal computer.  Software available online translates WiiMote movement during ankle flexion into control of online games.


Mechanisms considered for creating resistance to ankle flexion included springs, motors, magnets, and elastic bands.  Elastic bands were chosen because of their simplicity, low cost, wide resistance range, and fine resistance resolution.  Mechanisms considered for achieving a virtual reality interface included Microsoft Kinect, a custom designed system, and Nintendo WiiMote.  The WiiMote was chosen for its versatility, compatibility with existing software, and ability to provide quantitative data describing motion.


The Flex2Play hardware has four parts (Figure 1): footplate, base, hinge, and elastic bands.  The software consists of a WiiMote and a personal computer.

This image of the Flex2Play prototype shows a user’s foot fastened into the device with the four main components of the hardware labeled.
Figure 1. The Flex2Play Prototype

The aluminum footplate (Figure 2) has adjustable side plates and two adjustable velcro straps.  The side plates are secured with wing nuts and accommodate feet up to 12.5 cm wide.  The straps loop over the forefoot and around the ankle to secure the foot in place.  The forefoot strap can be set at one of three locations and accommodates feet up to 30cm long.

This CAD image of the Flex2Play footplate shows the adjustable side plates and the locations where the straps are connected to secure the foot.

Figure 2. The Footplate

The steel base of the Flex2Play (Figure 3) provides structural stability, prevents tipping during standard use, and accommodates users of various heights and chairs.  Height adjustment is facilitated by nested telescoping posts secured with a PTO lock pin.  Rubber feet prevent slipping and allow the Flex2Play to be used on many floor surfaces.

This CAD image of the base shows the elastic band connection points and the telescoping post that accommodates users of various heights.

Figure 3. The Base


The hinge of the Flex2Play (Figure 4) connects the footplate to the base and rotates about a single axle parallel to the ground to allow the user to dorsiflex and plantarflex.  Plastic bushings reduce friction in the hinge.

This CAD image of the hinge shows where the base connects to the footplate. It shows the plastic bushings that reduce friction in the hinge.

Figure 4. The Hinge


Elastic Bands
The Flex2Play is designed so that the inexpensive and widely available Thera-Band brand elastic bands (6) can be selected by a therapist during patient evaluation, tied securely to carabiners by the therapist, and then sent home for the user to clip onto the Flex2Play before training.  The therapist selects the band color and length for each patient based on the resistive force desired.

Virtual Reality Interface
A Nintendo WiiMote strapped to the footplate translates footplate motion into computer game control.  The WiiMote connects to a computer via Bluetooth and uses Glove Programmable Input Emulator (GlovePIE), a software program freely available online, to control the games (7).  Thus far, code has been written for Flex2Play use with three games (Tetris, 2-D Racing, Pong).  The code is adaptable so that each patient’s range of motion allows full game control.

The completed Flex2Play prototype provides resistance to dorsiflexion, plantarflexion, or both movements (Figure 5).

This diagram shows how the prototype can provide resistance to both dorsiflexion and plantarflexion at various levels of resistance through the user’s range of motion.

Figure 5. Flex2Play Training Capabilities

The Flex2Play prototype was verified and validated to ensure it met the design specifications.  Results are summarized in Table 1.

This table displays the design aims to be met by the prototype, the criterion for the device to meet the aims, and the results from verifying the actual prototype.

Table 1. Summary of Device Verification Results

Verification was successfully completed for all aims except one; we were unable to produce enough force to fully verify whether the Thera-band connection points could withstand the upper limits (originally chosen with a large safety factor).  Although the maximum forces were not tested, the device was confirmed to withstand more force than is experienced during standard use in the target population.


The Flex2Play prototype was validated during three stages of user-evaluation to determine whether the device fulfills user needs and withstands both intended and unintended uses.  The three stages of evaluation were completed under lab-approved protocol and included testing with a Flex2Play team member, healthy children (ages 5, 11, 19), and a young adult with cerebral palsy.

For all three sets of users, the sideplates and foot straps effectively secured the foot without causing discomfort.  The users could all move the Flex2Play through their full range of motion and the device did not tip.  The height adjustment worked well for the children, although for the full grown adults it was challenging to find chairs that were high enough to maintain proper knee-hip alignment.

The virtual reality interface worked well for all users.  Users appeared and reported to enjoy using the device to play video games.  The interface was successfully adapted for the individual with CP so that her limited range of ankle movement did not impair her ability to fully control the video games.  The device was also effectively adapted to focus on training certain aspects of ankle movements: dorsiflexion vs plantarflexion and range of motion vs speed of motion.

User testing revealed several areas for device improvement.  Most notably, the original design of connecting elastic bands to the device via carabiners introduced slack and thereby reduced the range of motion through which resistance was provided.  For device testing we tied the bands directly to the Flex2Play, but having patients tie the bands at home is not an ideal solution as it would introduce variability and decrease therapist regulation of in-home training.  We have redesigned the Flex2Play’s hooks for future prototypes so that therapists can tie bands in a loop and users can hook the pre-tied loops directly onto the device.

Other areas for future development of the Flex2Play include fatigue testing, minor design improvements, expanding the variety of online games available for play, developing a graphic user interface for modifying game control, and creating a progress-monitoring system that records device use and summarizes patient progress for therapists.  We also observed that as the children became absorbed in the games they inadvertently moved in their chairs and engaged the whole body during ankle flexion.  Developing some form of seat restraint may therefore help to ensure the ankle muscles are safely and efficiently trained during Flex2Play use.

The Flex2Play prototype was built for $297.23.  The next phase of developing the device is to demonstrate clinical efficacy through a controlled intervention study.  The estimated market size (based on the number of children in the US ages 5-18 with spastic CP) for the Flex2Play is 114,000.  Assuming a 1% per year market penetration, the device could therefore sell ~1000 devices per year.  The Flex2Play team has identified several ways to update the current design to significantly reduce materials and manufacturing costs if mass produced.  Despite these reductions, after the addition of regulatory, warranty, overhead, and other costs associated with bringing a device to market we anticipate that the shelf cost of the Flex2Play may be greater than our original $200 target.  For example, if materials and manufacturing costs are reduced to $100, the market price may come to three times that or roughly $300.  Fortunately, therapist consultations involving the Flex2Play may be eligible for reimbursement under CPT code 97110.  Once clinical efficacy is demonstrated the device itself may also be partially reimbursed by insurance companies.  These steps would lessen the financial burden to patients.

Furthermore, the feedback we received from users was encouraging and it appears that the Flex2Play has potential to expand into markets beyond children with CP.  In particular, the device may be useful for adults with CP, individuals in stroke rehabilitation, and athletes in sports rehabilitation.  Expanding the Flex2Play to these additional markets not only offers the potential for further cost reductions, but also offers hope for providing clinical benefits to a wider spectrum of patients through the use of this engaging in-home ankle strength trainer.


Funding for this project was provided by the Program in Occupational Therapy and the Department of Biomedical Engineering at Washington University.  We thank several individuals who have been crucial to the success of our project: Jack Engsberg, Joe Klaesner, and Frank Yin for their guidance; Bob Arcipowski and students at South Tech High School for their machining skills and eagerness to help; Jake Lohse and Michael Rogg for enabling our partnership with South Tech High School; Ben Harmon for CAD assistance; and the participants who have graciously donated their time to test the Flex2Play.


1. Wu, Y.-N., Ren, Y., Hwang, M., & Gaebler-Spira, D. (2010). Efficacy of Robotic Rehabilitation of Ankle Impairements in Children with Cerebral Palsy. 32nd Annual International Conference of the IEEE EMBS, 4481-4484.

2. Bax, M., Flodmark, O., & Tydeman, C. (2007). Definition and classification of cerebral palsy. From syndrome toward disease. Dev Med Child Neurol Suppl, 109, 39–41.

3. Aisen, M., Kerkovich, D., Mast, J., Mulroy, S., Wren, T., Kay, R., & Rethlefson, S. (2011). Cerebral Palsy: Clinical Care and Neurological Rehabilitation. Lancet Neurology, 10(9), 844-52.

4. Engsberg, J. R., Ross, S. A., Olree, K. S., & Park, T. S. (2000). Ankle spasticity and strength in children with spastic diplegic cerebral palsy. Developmental Medicine & Child Neurology, 42-47.

5. Dodd, K., Taylor, N., & Graham, H. (2003). A randomized clinical trial of strength training in young people with cerebral palsy. Developmental Medicine and Child Neurology, 652-657.

6. The Hygenic Corporation. (2006). Resistance Band & Tubing Instruction Manual. Akron, OH, USA.

7. Kenner, C. (2011). Ready to Download GlovePIE. Retrieved February 5, 2012, from http://glovepie.org/lpghjkwer.php


Kelly Hill, Amanda Meppelink, Elizabeth Phillips, Washington University in St. Louis


3507 Alba Place, Fairfax, VA 22031


, , , ,

Powered by WordPress. Designed by WooThemes

Skip to toolbar