Effectiveness of the RAPAEL Smart Board for Upper Limb Therapy inStroke Survivors: A Pilot Controlled Trial

Joonwoo Park1, Hee-Tae Jung2, Jean-Francois Daneault3,Sungji Park4, Taekyeong Ryu4, Yangsoo Kim4, and Sunghoon Ivan Lee2

Abstract— We aim to assess the effectiveness of using theRAPAEL Smart Board as an assistive tool for therapists inclinical rehabilitation therapy settings and to investigate if itcan be used to improve the motor recovery rate of strokesurvivors. The RAPAEL Smart Board is a therapy tool wheretherapists actively engage patients, giving necessary verbaland physical interventions as in traditional treatment sessions.We conducted a randomized controlled study with 17 strokesurvivors. An experimental group received therapy using theRAPAEL Smart Board for 30 minutes a day, 5 days per week,for 4 weeks in addition to their traditional treatments (i.e.30 minutes of functional arm movement therapy). A controlgroup received two 30-minute sessions of traditional treatment5 days per week, for 4 weeks. The upper-extremity functionwas measured using the Wolf Motor Function Test beforeand after the 4-week interventions. Our results demonstratethat using the RAPAEL Smart Board, in combination withtraditional treatment, significantly improves motor recoverywhen compared to traditional treatments alone.

I. INTRODUCTION

An increasing number of people suffer from strokes world-wide, many of whom are left with permanent disabilities.To improve motor recovery and restore normal functionalbehavior, stroke survivors need to undergo repetitive task-oriented therapeutic exercises [1]–[3]. However, it is chal-lenging for therapists to keep patients actively engaged inmundane repetitive movements during therapeutic exercises,which often leads to suboptimal motor recovery of patientsin daily clinical sessions.

To improve the level of patient engagement in rehabili-tation therapy, researchers have investigated the feasibilityof using games for stroke rehabilitation [4], [5]. It is envi-sioned that games will be entertaining, causing patients toparticipate more actively in repetitive exercises, which willin turn result in superior therapeutic outcomes compared totraditional rehabilitation interventions [6]–[8]. Commercialgaming devices, such as Nintendo Wii, Microsoft Kinect,and Sony PlayStation, have been the most widely studiedapparatus for this purpose due partially to their accessibility.

1J. Park is with Smilegreen Child Development Center, Daegu, S. Korea.This study was conducted while he was with College of RehabilitationSciences, Daegu University, S. Korea.

2H. Jung and S. I. Lee are with College of Information and ComputerSciences, University of Massachusetts Amherst, MA, USA.{hjung,silee}@cs.umass.edu

3J.-F. Daneault is with Department of Rehabilitation and MovementSciences, Rutgers University, NJ, USA.jf.daneault@rutgers.edu

4J. Jeong, T. Ryu and Y. Kim are with Heeyeon Rehabilitation Hospital,Changwon, S. Korea.{rtk,ysk}@silver4.net

Fig. 1. The RAPAEL Smart Board and an accompanying therapeutic game(Courtesy of Neofect Rehabilitation Solutions [11])

Although they have been shown to be effective, these deviceswere originally developed for entertainment for individualswith normal motor function [9], [10]. Subsequently, move-ments that are induced by these consoles and the accom-panying games may not address the therapeutically desirablemovements nor focus on the paretic limbs for stroke patients.In addition, the difficulty of the movements required by thosegames may be inappropriate for patients with wide range offunctional impairments. While the previous positive results ofusing video games lays a good foundation for future research,there is a need to investigate the effectiveness of utilizingthe device and the games, that are specifically designed forrehabilitation, in the actual routine clinical practice.

In the current study, we investigate the clinical effective-ness of using the RAPAEL Smart Board (Neofect, SouthKorea) [11], a game-based rehabilitation tool that has beenspecifically developed for the implementation of therapeuticarm exercises in the routine clinical setting. The results ofthe study suggest that rehabilitation therapies based on theRAPAEL Smart Board may induce better motor recovery ratewhen compared to traditional therapies of matching amountof therapy time.

II. RAPAEL SMART BOARD

The RAPAEL Smart Board is a rehabilitation solutionthat aims to improve arm function by practicing gravity-compensated movements (i.e., movements on the board)(Fig. 1). The patient grabs the handle that is placed abovethe board, instrumented with a location-based sensor thatcan recognize the handle’s horizontal location. Unlike other

(a) A game that requires horizontal point-to-point movements to feed selectedanimals

(b) A game that requires coordinated horizontal rotation movements to mixflour in a bowl and make dough

Fig. 2. Categories of games that focus on various aspects of arm movements (Courtesy of Neofect Rehabilitation Solutions [11])

TABLE IDEMOGRAPHIC AND CLINICAL INFORMATION ABOUT THE PARTICIPATING STROKE SURVIVORS

Experimental Group Control Group

Age Sex Affected side Diagnosis Chronicity Age Sex Affected Side Diagnosis Chronicity

77 Female Right Ischemic 15 75 Female Right Ischemic 1276 Female Right Ischemic 27 77 Female Left Ischemic 986 Female Right Ischemic 27 76 Female Right Ischemic 2650 Female Left Hemorrhagic 9 70 Female Left Hemorrhagic 1181 Female Left Hemorrhagic 6 80 Female Right Ischemic 1768 Female Right Hemorrhagic 24 80 Female Left Ischemic 968 Male Left Ischemic 8 73 Female Left Hemorrhagic 2665 Male Left Ischemic 14 92 Female Right Hemorrhagic 2751 Male Right Hemorrhagic 12 - - - - -

entertainment game consoles that are targeted for normallyfunctioning population, the RAPAEL Smart Board is de-signed to enable to customize a personalized exercise byselecting the game and constraining the degree of freedomin the handle to induce targeted movements in therapeuticallymore desirable movements [12]. The exercise games aredisplayed on the screen (Fig. 2). The Smart Board includesthree different categories of game. Games in category Afocus mostly on shoulder horizontal abduction/adduction andshoulder internal/external rotation. Games in category Bfocus mostly on elbow flexion/extension and shoulder flex-ion/extension. Games in category C focus on the combinationof the aforementioned movements. Since the Smart Boardcan track patients’ hand position but not their posture andmovement quality, it is recommended that therapy sessionsare supervised by experienced therapists.

III. METHODS

A. Study Population

The study was reviewed and approved by the InstitutionalReview Board. We recruited stroke survivors who had beenhospitalized in Heeyeon Rehabilitation Hospital, South Ko-rea. Patients who had a stroke at least 3 months before theenrollment were considered for participation to avoid theconfounding effect of spontaneous recovery. Patients with

significant comorbidity that could affect treatment outcomewere excluded. To ensure that patients could understandand follow the instructions, patients with a score of 18or below on the Mini Mental State Examination (MMSE)were excluded [13]. To avoid unexpected injury due tomuscle tone, patients with a score above 1+ on the ModifiedAshworth Scale (MAS) for the upper-limb being trainedwere also excluded [14]. Those who have visual neglectand other obvious issues in vision were not included in thestudy. Lastly, during the therapy sessions, therapists activelyengaged the patients by helping and correcting their armmovements and posture. Hence, patients experiencing painduring movements that spanned their passive as well as activerange of motion were excluded from the study.

B. Study Interventions

Study participants were randomly assigned to the experi-mental and control groups. The Smart Board therapy sessionswere administered by the patients’ regular therapists suchthat researchers were not involved in the therapy process.In order to minimize the effect from the initial adaptationperiod for patients and therapists to using the RAPAEL SmartBoard, the experimental group was exposed to the systemfor 30 minutes per day, five days per week, for a four-weekperiod prior to the initiation of the study. To normalize thetherapy dosage, the control group received two 30-minute

102 patients were assessed for eligibility

76 did not meet inclusion criteria

26 underwent randomization

13 were assigned to RAPAEL group 13 were assigned to conventional group

1 discharged

3 withdrew consent

2 discharged

3 withdrew consent

9 completed the study and were

included in analysis

8 completed the study and were

included in analysis

Fig. 3. Enrollment and outcomes

conventional therapy sessions five days a week, for fourweeks prior to the study.

After the 4 weeks of initial exposure, all participantsunderwent 4 weeks of treatment in the same manner. Par-ticipants from both groups conducted neurodevelopmentaltherapy, and interventions related to sensory stimulation andabnormal pattern suppression in their 30-minute conventionalsessions. The participants in the experimental group used theRAPAEL Smart Board during the second session. The gameswere chosen by the patients’ therapists in order to maximizethe effectiveness of the treatment and patient engagement.Some of the games that were frequently used during thetreatment included feeding animals and mixing flour dough(Fig. 2). The participants in the control group underwenttask-oriented therapeutic exercises, such as reaching to andmoving an item, that would match the movements facilitatedby the RAPAEL Smart Board exercises in their second ses-sion. Because the recovery rate in the routine clinical settingis the focus of the study, there was no particular number ofmovement repetitions that was enforced for neither groups.

C. Outcome Measures

Before and after the 4 weeks of treatment, the motorfunction of participants was assessed using the Wolf MotorFunction Test (WMFT) [15]. Assessments and treatmentswere done in a blind fashion. All assessments were doneby one trained clinician to minimize variability, who werenot involved in patient assignment nor treatment.

D. Statistical Analysis

A two-way mixed model analysis of variance (ANOVA)was used to identify differences between groups and theeffect of the intervention on the WMFT scores. Tukey’s posthoc test was used to assess significant differences. Thresholdfor significance was set to α < 0.05 a priori.

IV. RESULTS

Table II summarizes the WMFT scores from the exper-imental group and the control group at baseline and post-intervention. The two-way mixed model ANOVA demon-strated that there was a significant difference in condition(pre-post) (F = 34.81; p = 0.00003) as well as a signifi-cant interaction effect between group and condition (F =13.85; p = 0.002) but not significant group effect (F =2.04; p = 0.17). Since the goal of the current study wasto examine the effectiveness of the RAPAEL Smart Board,we then focused on specific statistical contrasts: was therea significant improvement in WMFT scores following thedifferent interventions and were there differences betweenthe experimental and control groups?

Firstly, Tukey’s post hoc test revealed that there was nosignificant difference between the experimental group andthe control group at baseline (q = 1.19; p = 0.41). Similarly,Tukey’s post hoc test revealed that there was no significantdifference between the experimental group and the controlgroup after the intervention (q = 1.66; p = 0.26)

Secondly, Tukey’s post hoc test revealed that there was nosignificant difference between baseline and post-interventionWMFT scores in the control group (q = 2.20; p = 0.14).However, Tukey’s post hoc test revealed that the improve-ment in WMFT score between baseline and post-interventionin the experimental group was statistically significant (q =9.83; p = 0.0002).

V. DISCUSSION

In this study, the effectiveness of the RAPAEL SmartBoard was investigated. To minimize the initial learningcurve and initial excitement effects by both patients and ther-apists on the exercise quality, the experiment group was ex-posed to the experiment treatment for a 1-month period priorto the actual experiment. Our study results, derived fromdata obtained from 17 post-stroke patients, demonstratedthat patients who underwent a combination of one RAPAEL

TABLE IIDESCRIPTIVE STATISTICS OF THE WOLF MOTOR FUNCTION TEST (WMFT) SCORES FOR BOTH GROUPS

Experimental Group Control Group

Condition N Mean Median StDev SEM Condition N Mean Median StDev SEM

Pre 9 49.11 56.00 20.85 6.95 Pre 8 36.88 44.50 21.28 7.52Post 9 55.44 60.00 21.26 7.09 Post 8 38.38 44.50 21.17 7.48

Smart Board therapy session and one traditional therapysession achieved statistically significant improvement in theirarm function measured by the WMFT when compared tothose who underwent two traditional therapy sessions; bothgroups received the interventions for five days a week fora one-month period. This demonstrates that the RAPAELSmart Board is a suitable tool that can help therapists treatpost-stroke patients in the routine clinical setting, whichmay lead to superior recovery rate and therapeutic outcomesin the same given therapy hours. Since the Smart Boardcannot capture the finger/hand/wrist movements, therapistsmay complement this by additionally employing the SmartGlove sessions [16].

However, we would want to be cautious when it comes toidentifying the fundamental cause of the observed differencein the arm recovery rate. Patients’ behavior observed inthe video recording did not necessarily speak for betterengagement to exercise therapy during the Smart Boardsessions. This was corroborated by the passing remarks fromthe therapists, who participated in our study, that presentingthe game alone was not enough increase the patients’ en-gagement significantly. Further research appears necessaryto understand the effects of games to physical therapies inmore depth.

Limitations of this study include a small number ofsubjects and a preponderance of female subjects (82.35%).A future study may recruit a larger number gender-balancedsubjects to establish more generalized and conclusive resultsregarding the clinical effectiveness. Also, the current experi-ment results do not capture the therapeutic effect that mighthave occurred in the familiarization phase.

VI. CONCLUSION

We found that there were trends toward greater improve-ment for post-stroke patients with upper-limb impairmentwhen Smart Board-assisted therapy was provided along withconventional therapy, compared to when patients received thematching therapy time of conventional therapies only. Ourpromising results lay the foundation for larger-scale studiesto examine the use of video games and sensors in strokerehabilitation therapies.

ACKNOWLEDGMENT

The authors would like to thank the staff members fromHeeyeon Rehabilitation Hospital, S. Korea, for their sup-port throughout the study. The authors declare no financialconflicts of interest. Specifically, none of the authors are

employed by nor received funding from Neofect Rehabil-itation Solutions (S. Korea). The authors appreciate KallieWhritenour for her help in proofreading the manuscript.

REFERENCES

[1] B. French, L. H. Thomas, M. J. Leathley, C. J. Sutton, J. McAdam,A. Forster, P. Langhorne, C. I. M. Price, A. Walker, and C. L. Watkins,“Repetitive task training for improving functional ability after stroke,”Stroke, vol. 40, no. 4, pp. e98–e99, 2009.

[2] B. M. Quaney, J. He, G. Timberlake, K. Dodd, and C. Carr, “Vi-suomotor training improves stroke-related ipsilesional upper extremityimpairments,” Neurorehabilitation and Neural Repair, vol. 24, no. 1,pp. 52–61, 2010.

[3] J. M. Veerbeek, E. Wegen, R. Peppen, P. Wees, E. Hendriks, M. Ri-etberg, and G. Kwakkel, “What is the evidence for physical therapypoststroke? a systematic review and meta- analysis.” PloS One, vol. 9,no. 2, p. e87987, 2014.

[4] G. Saposnik and M. Levin and Stroke Outcome Research Canada(SORCan) Working Group, “Virtual reality in stroke rehabilitation:a meta-analysis and implications for clinicians,” Stroke, vol. 42, no. 5,pp. 1380–1386, 2011.

[5] K. Thomson, A. Pollock, C. Bugge, and M. Brady, “Commercialgaming devices for stroke upper limb rehabilitation: a systematicreview,” International Journal of Stroke, vol. 9, no. 4, pp. 479–488,2014.

[6] M. K. Holden and T. Dyar, “Virtual environment training: A newtool for neurorehabilitation,” Journal of Neurologic Physical Therapy,vol. 26, no. 2, pp. 62–71, 2002.

[7] K. Laver, S. George, S. Thomas, J. E. Deutsch, and M. Crotty, “Virtualreality for stroke rehabilitation,” Stroke, vol. 43, pp. e20–e21, 2012.

[8] G. N. Lewis and J. A. Rosie, “Virtual reality games for movementrehabilitation in neurological conditions: how do we meet the needsand expectations of the users?” Disability and Rehabilitation, vol. 34,no. 22, pp. 1880–1886, 2012.

[9] S. Flynn, P. Palma, and A. Bender, “Feasibility of using the sonyplaystation 2 gaming platform for an individual poststroke: A casereport,” Journal of Neurological Physical Therapy, vol. 31, pp. 180–189, 2007.

[10] G. Yauzer, A. Senel, M. B. Atay, and H. J. Stam, ““playstation eyetoygames” improve upper extremity-related motor functioning in subacutestroke: A randomized controlled clinical trial,” European Journal ofPhysical and Rehabilitation Medicine, vol. 44, no. 3, pp. 237–244,2008.

[11] NEOFECT Rehabilitation Solutions, “RAPAEL Smart Board — NEO-FECT,” http://www.neofect.com/en/product/boards, January 28 2018.

[12] M. F. Levin, J. A. Kleim, and S. L. Wolf, “What do motor “recovery”and “compensation” mean in patients following stroke?” NeurorehabilNeural Repair, vol. 23, pp. 313–319, 2009.

[13] M. F. Folstein and S. E. Folstein, ““Mini-Mental State” a practicalmethod for grading the cognitive state of patients for the clinician,”Journal of Psychiatric Research, vol. 12, no. 3, pp. 189–198, 1975.

[14] C. P. Charalambous, “Interrater reliability of a modified ashworthscale of muscle spasticity,” in Classic Papers in Orthopaedics, P. Ba-naszkiewicz and D. Kader, Eds. London: Springer, 2014, pp. 415–417.

[15] S. L. Wolf, P. A. Catlin, M. Ellis, A. L. Archer, B. Morgan, andA. Piacentino, “Assessing wolf motor function test as outcome measurefor research in patients after stroke,” Stroke, vol. 32, pp. 1635–1639,2001.

[16] H. Jung, H. Kim, J. Jeong, B. Jeon, T. Ryu, and Y. Kim, “Feasibilityof using the rapael smart glove in upper limb physical therapy forpatients after stroke: A randomized controlled trial,” in Proceedingsof the 39th Annual International Conference of the IEEE Engineeringin Medicine and Biology Society, 2017.