Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Game-Based Rehabilitation for Myoelectric Prosthesis Control

Game-Based Rehabilitation for Myoelectric Prosthesis Control Background: A high number of upper extremity myoelectric prosthesis users abandon their devices due to difficulties in prosthesis control and lack of motivation to train in absence of a physiotherapist. Virtual training systems, in the form of video games, provide patients with an entertaining and intuitive method for improved muscle coordination and improved overall control. Complementary to established rehabilitation protocols, it is highly beneficial for this virtual training process to start even before receiving the final prosthesis, and to be continued at home for as long as needed. Objective: The aim of this study is to evaluate (1) the short-term effects of a commercially available electromyographic (EMG) system on controllability after a simple video game-based rehabilitation protocol, and (2) different input methods, control mechanisms, and games. Methods: Eleven able-bodied participants with no prior experience in EMG control took part in this study. Participants were asked to perform a surface EMG test evaluating their provisional maximum muscle contraction, fine accuracy and isolation of electrode activation, and endurance control over at least 300 seconds. These assessments were carried out (1) in a Pregaming session before interacting with three EMG-controlled computer games, (2) in a Postgaming session after playing the games, and (3) in a Follow-Up session two days after the gaming protocol to evaluate short-term retention rate. After each game, participants were given a user evaluation survey for the assessment of the games and their input mechanisms. Participants also received a questionnaire regarding their intrinsic motivation (Intrinsic Motivation Inventory) at the end of the last game. Results: Results showed a significant improvement in fine accuracy electrode activation (P<.01), electrode separation (P=.02), and endurance control (P<.01) from Pregaming EMG assessments to the Follow-Up measurement. The deviation around the EMG goal value diminished and the opposing electrode was activated less frequently. Participants had the most fun playing the games when collecting items and facing challenging game play. Conclusions: Most upper limb amputees use a 2-channel myoelectric prosthesis control. This study demonstrates that this control can be effectively trained by employing a video game-based rehabilitation protocol. (JMIR Serious Games 2017;5(1):e3) doi: 10.2196/games.6026 https://games.jmir.org/2017/1/e3/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e3 | p. 1 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Prahm et al KEYWORDS upper limb prosthesis control; upper extremity amputees; gaming; serious games; neuromuscular rehabilitation; intrinsic motivation; EMG control This game is based on rhythm and speed, and requires a fast Introduction reaction from the player and an immediate transmission of the processed EMG signals to the gaming system. Similarly, a The initial control of a myoelectric prosthesis can be a rehabilitation concept for stroke patients using a modified frustrating experience, especially after the already traumatic version of the WiiMote control is used for rehab purposes of event of losing a limb. Due to the nonintuitive interface, which upper limb amputees, in which EMG signals are matched to the handles a complex mechatronic system, the cognitive demand keys of the WiiMote [22]. However, controls are only limited for controlling the prosthesis is high and further delays the actual to two motions. Other groups chose a game similar to the arcade use of the device in everyday life [1,2]. At least 50% of upper classic Pong, in which the user’s muscle activity is mapped into extremity amputees report problems with prosthesis control and a paddle motion that hits a ball into their opponent’s court [23]. functionality [3,4], which can be attributed to the need for Although those approaches can be motivating, the necessary receiving more training in handling the prosthesis [5,6]. By actions are not very intuitive and are not directly transferable providing more training opportunities, the user can fully benefit to the handling of a prosthesis [24]. from the technical functions of the prosthesis. This study presents an interface between a computer and a To prepare the residual muscles and induce specific brain commercially available surface EMG electrode system plasticity, having access to a prosthesis itself is not necessarily (Ottobock Healthcare GmbH, 13E200), which is commonly needed. Regaining muscle strength and coordination is a used for controlling prostheses, to evaluate the short-term effects cognitively exhausting and repetitive process, during which the on controllability after a video game-based rehabilitation proper execution of movements is reestablished using surface protocol. electromyographic (EMG) feedback [2,7,8]. Through physiotherapy, patients are presented with a variety of tasks This study, compared to previous studies, prompted participants promoting the development of coping strategies for dealing with to not only conduct repetitive agonist and antagonist muscle the activities of daily living, and introducing the embodiment activation, but also to train and exert sustained contractions over of the prosthetic system itself. To effectively control their a short period of time, perform precisely timed contractions, prosthesis, patients need to learn how to properly contract their and elicit simultaneous contraction of both muscles and muscle muscles; the strength of activation and isolation of a single groups. These functions are similar to how patients would muscle are important parameters [1,9,10]. control a real prosthesis as they interact with their environment. Although the standard rehabilitation program offers direct Methods functional benefits, its main shortcomings are the lack of motivation for patients to pursue it without the involvement of Eleven able-bodied participants, who had no prior experience a therapist throughout the lengthy process. In addition to the in EMG control, took part in appraising the benefits of the video loss of functionality, patients may suffer from posttraumatic game-based training. The categories of the EMG controllability depression, further decreasing motivation for rehabilitation [11]. assessments that were evaluated consisted of a provisional Transferring traditional EMG rehabilitation protocols to a virtual maximum voluntary muscle contraction for calibration, precision setting, and incorporating video games into the training process, control, electrode separation, and endurance control by retracing can potentially increase the patient’s engagement and a sine curve with the EMG signal. These assessments are further perseverance [12]. This approach also provides medical explained in detail in the Electromyographic Assessments professionals with quantitative data of the patient’s performance. subsection. Three video games and their respective control Many studies report that the progress achieved during variations were evaluated for their motivational factors and rehabilitation based on a playful concept is faster and superior feasibility. Two questionnaires were given to evaluate (1) the to conservative physiotherapeutic exercises [9,13-15]. These video games and the input method, and (2) intrinsic motivation. rehabilitation games are especially popular in older adults [16], This study was approved by the ethics committee at the Medical and when treating patients affected by stroke [17,18] and University of Vienna (number 1301/2015) and all study Parkinson’s disease [19,20]. Various research groups have participants read and signed the consent form before taking part. addressed adding virtual games to an otherwise dull routine in Participants the area of upper limb amputee rehabilitation. There is, however, Eleven naïve, able-bodied participants without any known a difference between virtual or augmented reality environments neurological or muscular impairments participated in the study. and using commercially available video games during All participants had normal or corrected-to-normal vision and therapeutic interventions [9,21]. The latter provides greater were instructed and accompanied by the examiner throughout accessibility and allows patients to easily set up the games at the entirety of the study. home, and games can be chosen that are proven to motivate the players and maintain engagement over a longer period of time [15]. An example of a commercially available video game that has been interfaced using EMG signals is Guitar Hero [14]. https://games.jmir.org/2017/1/e3/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e3 | p. 2 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Prahm et al Participants were invited two times and had three test sessions Experimental Protocol in total: one Pregaming and Postgaming measurement, both Each participant was seated comfortably in front of two conducted on the same day; and one Follow-Up measurement computer screens. One screen displayed the acquired EMG data to evaluate short-term retention rate two days later. per electrode channel, the other screen showed the game that the participant was playing. Two active surface EMG electrodes Participants were initially instructed to perform three basic EMG (Ottobock Healthcare GmbH 13E200) were positioned on top assessments: the provisional maximum voluntary contraction of the prominent flexor and extensor muscles of the wrist on (MVC) level, accuracy of electrode control, and muscle the participant’s nondominant side (see Figure 1). This was endurance. After a short break, participants were presented with done to match the handedness of the amputees, which is always three computer games in randomized order. After each transferred on the intact limb regardless of the preimpairment EMG-controlled game, they were asked to complete a short state. Amplification and electrode placement remained the same user evaluation survey regarding the gaming experience. After throughout the sessions. Each electrode delivered root mean the third and final game, a modified questionnaire aiming at square (RMS) at the 100-hertz rate of the recorded EMG signal intrinsic motivation (Intrinsic Motivation Inventory; IMI) was following the embedded filtering and rectification. completed. https://games.jmir.org/2017/1/e3/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e3 | p. 3 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Prahm et al Figure 1. The experimental set-up. The outcome measures that were considered included Electromyographic Assessments fluctuations of RMS EMG signals over expected EMG signals. To investigate the changes in overall controllability, three basic Both electrodes typically show some negligible offset activation assessments evaluating approximate strength, muscle precision (due to common noise) during the idle state of the forearm control and coordination, and muscle endurance were performed. muscles (see Figure 2). https://games.jmir.org/2017/1/e3/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e3 | p. 4 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Prahm et al Figure 2. The interface for the precision control assessment and separation of electrode activation. Electrode 1 is used to assess precision control while electrode 2 indicates either the separation or cocontraction of both electrodes. a) The red triangle indicates the goal activation level that participants must reach with their electromyographic signals. b) The black bar marks the threshold at which the electrode is considered active. If the electromyographic signal passed this bar, electrode separation failed and cocontraction is detected. concept of reaching a certain threshold for activating the Maximum Voluntary Contraction Test electrode corresponds to the actual execution of prosthetic The MVC test was used as a calibration of the voltage detected movements [1,25]. The outcome measure was the binary by the electrodes, and assessed the MVC force (averaged over activation of the opposing electrode and the overall percentage 3 trials) for each of the two electrode channels. Participants of activation of the opposing electrode per participant. were asked to maximally contract one muscle and to hold this Assessment of Endurance Control contraction for 1.3 seconds, of which only the last second was taken for calculating the activation baseline. The Assessment of Endurance Control test assessed muscle coordination and muscle fatigue while the participants used Assessment of Precision Control their EMG signals to closely follow a sine curve (1/4 hertz) on The Assessment of Precision Control test evaluated the the screen until they felt fatigued. The estimated force needed participant’s fine EMG control accuracy. The range of this test to reach the peaks of the sine curve corresponds to 60% MVC. was adapted based on the outcome of the MVC test. For each A positive value corresponded to activation of the first electrode, electrode, the participant was asked to reach 30 randomly while a negative value corresponded to the second. Electrode preselected activation levels in the range of 10-90% MVC, and activation needed to be separate to reach the peaks of the sine sustain them for 300 milliseconds each. The required level of curve. The minimum time to be reached in this test was 15 activation was indicated by a triangular mark on the EMG bar minutes. The outcome measure was the EMG signal deviation (see Figure 2 a). A total of 30 marks (3 trials consisting of 10 from the desired sine curve, given as correlation r² [26]. levels) were performed for each electrode. The percentile Games deviation from the mark was taken as outcome measure. Randomization of the goal activation marks took place once Three different open-source games were used in this study: a before the beginning of the study and was kept constant between racing game, a dexterity game, and a rhythm-based game. Each all participants. game featured its own respective control method (see Figure 3). The general input mechanism was to substitute keyboard Assessment of Separation events with EMG activation. Participants controlled two The Assessment of Separation test is a subsection of the electrodes, making two concurrent keys possible at a time. Those Assessment of Precision Control, and determined whether motions represented one degree of freedom (DoF). However, participants could separately control one muscle or if the to allow for more than just one DoF steering, participants could opposing electrode was activated (cocontraction) during the perform a cocontraction (a quick simultaneous activation of tasks of the Precision Control assessment. Depending on the both opposing muscles) and switch to a second DoF. For games MVC, a threshold was set that corresponded to the point of that did not offer the option of cocontraction to switch through EMG activation at which an electrode was considered active keys, the dominant limb supported the control through direct (see Figure 2 b). This threshold was set at 15% MVC. The keyboard input. https://games.jmir.org/2017/1/e3/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e3 | p. 5 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Prahm et al Figure 3. An overview of the three games played by the participants, with their respective input methods and muscle contraction types needed to drive the game. EMG: electromyographic. across the screen. Required EMG activations were quick Racing Game contractions, sustained contractions over a certain time period, In the 3-dimensional racing game Super Tux Kart [27] the player and cocontractions to simultaneously activate two buttons. raced against the clock and computer-controlled adversaries. Questionnaires The participant controlled left and right turns solely with EMG signals, whereas accelerating and braking were controlled using Participants were given two questionnaires to complete: (1) a keyboard inputs with their dominant hand. Required EMG modified IMI questionnaire; and (2) a user evaluation survey activations were quick contractions and sustained contractions. about the EMG assessment, the games that were played, and control methods. Dexterity Game Modified Intrinsic Motivation Inventory Questionnaire In the dexterity game Pospos [28] the player had to maneuver through a 2-dimensional labyrinth and collect items. This game A modified 28-item version of the IMI [30-32] consisting of was controlled entirely through the participant’s EMG signals. five subscales was used to evaluate the experience with the Switching between the DoFs was done by cocontraction, which video games that were played. The five subscales formed scores corresponded to controlling the horizontal and vertical axes of for enjoyment, perceived competence, perceived choice, pressure the player. Required EMG activations were quick contractions, felt, and immersion. An additional six questions were added to sustained contractions, and cocontractions to switch between the last subsection to evaluate immersion into the games. The DoFs. questionnaire was adapted to the study by changing the words, “working” and, “doing” to, “playing”. The questionnaire Rhythm Game included statements such as, “I found the games very In the rhythm-based game Step Mania 5 [29] participants were interesting” and, “I felt tense while playing.” The statements prompted to activate 2 different arrow-shaped buttons using were rated on a 7-point Likert rating scale ranging from 1 (no, their EMG signal. The arrows had to be quickly pressed or held not at all) to 7 (yes, definitely). in time matching the rhythm of the note patterns that scrolled https://games.jmir.org/2017/1/e3/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e3 | p. 6 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Prahm et al Improvement was tested for significance with a related samples User Evaluation Survey Wilcoxon signed rank test. The user evaluation survey consisted of (1) rating the games that were played, (2) rating the input and player control methods Assessment of Endurance Control (see Figure 3), (3) rating the EMG assessment, and (4) The conformity of retracing the sine curve as a correlation r² identifying engaging elements within each game. This short was computed for defined time windows consisting of 30 survey about gaming experience was presented after every game seconds. The highest r² value was taken from a period of at least and included questions about the gameplay, fun factor, 200 seconds. A related samples Wilcoxon signed rank test motivation, and input and control methods. (Cronbach alpha=.05) examined the Pregaming measurement correlation with the Follow-Up measurement correlation. Statistical Analyses All analyses were conducted using IBM SPSS 20 and Matlab Questionnaires 2013b. Nonparametric tests were performed on data not meeting Intrinsic Motivation Inventory Questionnaire the requirement for normal distributions. Normal distributions Participants had to rank the 28 statements from 1 to 7, where 1 were assessed via graphical interpretation showing normal Q-Q represented I do not agree and 7 represented I agree. The plots, and with Shapiro-Wilk tests for normality as a numerical statements belonged to one of five categories and the ranking assessment. Significance was set at Cronbach alpha=.05. was averaged. An independent samples Mann Whitney U test Controllability was performed to describe the data. All categories except pressure had a high desirable rank. Maximum Voluntary Contraction This test was used to relatively set the maximum contraction User Evaluation Survey limit for the subsequent EMG tests, and was given as average This survey consisted of ranked statements on a 5-point scale, of the RMS electrode activation. MVC was measured three and multiple choice questions regarding game experience and times: before playing the games, directly after playing the preferences that were evaluated via a frequency analysis. games, and before the Follow-Up measurement. Results Assessment of Precision Control The outcome measure for this test was the percent deviation Controllability from the 30 goal points per electrode channel. The deviation from the goal was calculated in absolute values and set in Maximum Voluntary Contraction relation to the achieved MVC to derive a percent value of The MVC was used only as a calibration for the electrode deviation for each goal point. All 30 data points were treated channel voltage. However, it could be observed that the RMS as if they were performed consecutively. The goal points were values for the MVC test directly after the games showed an divided into three equidistant intensity sections ranging from increase instead of the expected decrease. Moreover, the same 10% MVC to 90% MVC. The Shapiro-Wilk test confirmed a can be observed for the Follow-Up session. normal distribution with P<.001. The mean and standard Assessment of Precision Control deviation were compared for significance for the activation levels (low, middle, high) and for measurement sessions Results showed a significant improvement in fine accuracy and (Pregaming, Postgaming, Follow-Up) with a Bonferroni electrode coordination from the Pregaming measurement to the corrected paired samples t-test. Follow-Up session (P=.001). Percentile deviation from the goal value was high and heterogeneous in the first two measurements Assessment of Separation (Pregaming and Postgaming), but significantly lower and more Threshold crossings were given in percentages for each of the homogeneous in the Follow-Up measurement (P=.001; see 3 goal activation levels over the 3 measurement sessions. The Figure 4). In all 3 measurement sessions, it was significantly three equidistant intensity levels ranged from 10-90%. harder to reach a high goal activation level compared to a low one (P=.002). https://games.jmir.org/2017/1/e3/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e3 | p. 7 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Prahm et al Figure 4. Development of the deviation of the electrode activation around the goal value levels (separated into low, middle, and high goal values) through all three measurement sessions (Pregaming, Postgaming and Follow-Up). MVC: maximum voluntary contraction. opposing electrode activation from the first to the last Assessment of Separation measurement session for low (P=.04) and middle (P=.02) goal Significantly better performance could be observed during low activation levels; however, this was not true for high intensity level electrode activation tasks compared to high goal activation goal activation (see Figure 5 b). There was a significant tasks within Pregaming (P=.02) and the Follow-Up session improvement in electrode separation overall, considering the (P=.04; see Figure 5 a). There was a significant decrease in first and last measurement sessions (P=.02). Figure 5. Opposing electrode activations over three measurement sessions (Pregaming, Postgaming and Follow-Up) and three goal activation areas, which the participants had to reach with their electromyographic signal (low, middle, and high). a) Comparison within the session. b) Comparison between the sessions. https://games.jmir.org/2017/1/e3/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e3 | p. 8 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Prahm et al shifted by half a period of the retraced sine curve. Nevertheless, Assessment of Endurance Control the Related Samples Wilcoxon Signed Rank Test determined Participants showed an improvement in muscle endurance a significant difference (P=.004) in r² performance between the control (see Figure 6). This result was true for all but one Pregaming and Follow-Up measurement sessions. participant, whose EMG activation expressed an offset and was Figure 6. Scores of the endurance assessment and comparison of Pregaming r² value to the Follow-Up value. High r² corresponds to close electromyographic retracing of the given sine curve. Participants enjoyed playing the games and felt immersed when Questionnaires doing so. Participants perceived playing the games as their own choice and felt competent and at ease while doing so. Pressure Modified Intrinsic Motivation Inventory was the only category in which a low rank was desirable. Results obtained from the IMI questionnaire can be viewed as mean and standard deviation of the five categories in Figure 7. https://games.jmir.org/2017/1/e3/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e3 | p. 9 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Prahm et al Figure 7. Means and standard deviations for 5 subscales of the modified intrinsic motivation questionnaire on a scale from 1 (low) to 7 (high). graphics rated last. Although participants claimed to prefer User Evaluation Survey 3-dimensional graphics, this finding did not reflect their rating Participants were asked to answer five questions for each game of what they enjoyed most in the games. The most motivating (Q1-Q5; see Figure 8). According to the user evaluation survey, aspects of games were (1) the gameplay, (2) to see one’s own the favorite game (derived from the score of Q1 and Q2) was high score, and (3) to clear upgrades. the racing game Super Tux Cart followed by the rhythm game Additionally, participants had to rate the EMG assessments after Step Mania 5. According to Q3 and Q4, participants preferred each session. Participants were asked about how important they to control the games with EMG signals only and to perform thought the EMG assessments were, and to rate the fun they different contraction lengths as well as cocontractions. In terms had while doing them. As can be seen in Figure 9, rating of the of motivation (Q5), the Pospos dexterity game ranked far behind importance of the EMG assessment increased until the the racing and rhythm games, which were equally well received. Follow-Up measurement (however, not significantly), while The most important components to ensure continued play and the participants enjoyed them significantly less (P=.002). enjoyment of a game were (listed according to importance): (1) Interestingly, a slight rise in rating the fun factor was observed the EMG control method, (2) the level of difficulty, (3) dynamic after the Follow-Up session. movements, and (4) collecting items. Music, atmosphere, and Figure 8. Mean and standard deviation ratings of the three games played, according to the survey that participants had to fill in after each game. https://games.jmir.org/2017/1/e3/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e3 | p. 10 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Prahm et al Figure 9. Mean and standard deviation for ratings of the electromyographic assessment after each session, ranging from 5 (very important/fun) to 1 (very unimportant/boring). EMG: electromyographic. the listed EMG tests. However, prolonged exposure to such Discussion stimuli would certainly lead to a loss of interest, which is sure to be maintained by the appealing context of a video game [15]. Results from this study demonstrate improvements in fine accuracy electrode activation and electrode separation from Limitations Pregaming EMG assessments to the Follow-Up measurements. The transferability of the obtained results to the amputee Surprisingly, the MVC values used as a baseline calibration population might be questioned, since this study was conducted also showed an increase, instead of the expected decrease, after strictly with healthy participants. However, based on the playing the games. This result could be due to either warmth outcomes reported in other myocontrol-based studies [34,35], or sweat that would influence the electrode resistance. it is reasonable to expect that the patient group would perform Additionally, this result is a strong indicator that the gaming similarly. session was not fatiguing for the participants. Performance during the precision control assessment, however, declined after This study was a short-term intervention, and can be viewed as playing the games. If, based on previous investigations, we a proof of concept. Further research will incorporate a long-term exclude fatigue, it is reasonable to assume that participants evaluation of video game-based interventions, as well as started losing their concentration by the end of the sessions. In additional exploration of advanced control mechanisms, such the Follow-Up measurement, a clear improvement in as those based on machine learning approaches [36-38]. performance was observed, which can be attributed to the Conclusion restoration of full focus combined with the obtained experience Most upper limb amputees use a 2-channel myoelectric from the previous session. prosthesis control. This study demonstrates that this control can Compared to previous studies [22-24,33], participants not only be effectively trained by employing a video game-based conducted repetitive flexor and extensor muscle activation, but rehabilitation protocol. Participants significantly improved their also sustained contractions over varying periods of time, electrode separation and fine muscle control. It could be shown performed precisely timed contractions, and executed that the enjoyment of the games was greater than that of the simultaneous contractions of both muscle groups. These actions EMG assessments, which decreased over time. Additionally, are similar to how patients would control a real prosthesis. engaging elements within each game could be identified. A subsequent study with an amputee population will show if the The motivational aspects of training gamification are clear, and information gained from healthy participants can be transferred are likely the main advantage compared to conventional to patients. The final outcome would be a robust system that techniques. It is reasonable to assume that certain improvement patients can operate outside of a clinical environment. of the EMG control could be observed by sole application of https://games.jmir.org/2017/1/e3/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e3 | p. 11 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Prahm et al Acknowledgments This work was supported by the Christian Doppler Research Foundation of the Austrian Federal Ministry of Science, Research and Economy, and by the European Research Council Advanced Grant DEMOVE (contract #267888). The authors would like to thank Korbinian Eckstein, MSc for system implementations. Authors CP and IV designed and conducted the study, led the analysis, and interpreted the data; all authors contributed to drafting and reviewing the manuscript. Conflicts of Interest None declared. References 1. Roche AD, Rehbaum H, Farina D, Aszmann OC. Prosthetic myoelectric control strategies: a clinical perspective. Curr Surg Rep 2014 Jan 25;2(3). [doi: 10.1007/s40137-013-0044-8] 2. Sturma A, Göbel P, Herceg M, Gee N, Roche A, Fialka-Moser V, et al. Advanced rehabilitation for amputees after selective nerve transfers: EMG-guided training and testing. In: Replace, Repair, Restore, Relieve – Bridging Clinical and Engineering Solutions in Neurorehabilitation. New York: Springer; 2014:169-177. 3. Kyberd PJ, Beard DJ, Davey JJ, Morrison JD. A survey of upper-limb prosthesis users in Oxfordshire. JPO 1998;10(4):84-91. [doi: 10.1097/00008526-199801040-00004] 4. Dudkiewicz I, Gabrielov R, Seiv-Ner I, Zelig G, Heim M. Evaluation of prosthetic usage in upper limb amputees. Disabil Rehabil 2004 Jan 7;26(1):60-63. [doi: 10.1080/09638280410001645094] [Medline: 14660200] 5. Weeks DL, Anderson DI, Wallace SA. The role of variability in practice structure when learning to use an upper-extremity prosthesis. JPO 2003;15(3):84-92. [doi: 10.1097/00008526-200307000-00006] 6. Roche AD, Vujaklija I, Amsüss S, Sturma A, Göbel P, Farina D, et al. A structured rehabilitation protocol for improved multifunctional prosthetic control: a case study. J Vis Exp 2015 Nov 06(105):e52968. [doi: 10.3791/52968] [Medline: 26575620] 7. Smurr LM, Gulick K, Yancosek K, Ganz O. Managing the upper extremity amputee: a protocol for success. J Hand Ther 2008 Apr;21(2):160-75; quiz 176. [doi: 10.1197/j.jht.2007.09.006] [Medline: 18436138] 8. Aszmann OC, Roche AD, Salminger S, Paternostro-Sluga T, Herceg M, Sturma A, et al. Bionic reconstruction to restore hand function after brachial plexus injury: a case series of three patients. The Lancet 2015 May;385(9983):2183-2189. [doi: 10.1016/s0140-6736(14)61776-1] 9. Anderson F, Bischof W. Augmented reality improves myoelectric prosthesis training. Int J Disabil Hum Dev 2014;13(3):349-354. [doi: 10.1515/ijdhd-2014-0327] 10. Dawson MR, Carey JP, Fahimi F. Myoelectric training systems. Expert Rev Med Devices 2011 Sep;8(5):581-589. [doi: 10.1586/erd.11.23] [Medline: 22026623] 11. Kotila M, Numminen H, Waltimo O, Kaste M. Depression after stroke: results of the FINNSTROKE Study. Stroke 1998 Feb;29(2):368-372 [FREE Full text] [Medline: 9472876] 12. Burke JW, McNeill MDJ, Charles DK, Morrow PJ, Crosbie JH, McDonough SM. Optimising engagement for stroke rehabilitation using serious games. Vis Comput 2009 Aug 27;25(12):1085-1099. [doi: 10.1007/s00371-009-0387-4] 13. Tatla SK, Shirzad N, Lohse KR, Virji-Babul N, Hoens AM, Holsti L, et al. Therapists' perceptions of social media and video game technologies in upper limb rehabilitation. JMIR Serious Games 2015;3(1):e2 [FREE Full text] [doi: 10.2196/games.3401] [Medline: 25759148] 14. Armiger R, Vogelstein R. Air-Guitar Hero: a real-time video game interface for training and evaluation of dexterous upper-extremity neuroprosthetic control algorithms. In: Biomedical Circuits and Systems Conference. 2008 Presented at: BioCAS; Nov 2008; IEEE p. 121-124. [doi: 10.1109/biocas.2008.4696889] 15. Lohse K, Shirzad N, Verster A, Hodges N, Van der Loos HF. Video games and rehabilitation: using design principles to enhance engagement in physical therapy. J Neurol Phys Ther 2013 Dec;37(4):166-175. [doi: 10.1097/NPT.0000000000000017] [Medline: 24232363] 16. van Diest M, Stegenga J, Wörtche HJ, Verkerke GJ, Postema K, Lamoth CJ. Exergames for unsupervised balance training at home: a pilot study in healthy older adults. Gait Posture 2016 Feb;44:161-167. [doi: 10.1016/j.gaitpost.2015.11.019] [Medline: 27004651] 17. Duncan PW, Horner RD, Reker DM, Samsa GP, Hoenig H, Hamilton B, et al. Adherence to postacute rehabilitation guidelines is associated with functional recovery in stroke. Stroke 2002 Jan;33(1):167-177 [FREE Full text] [Medline: 11779907] 18. Lang CE, Macdonald JR, Reisman DS, Boyd L, Jacobson KT, Schindler-Ivens SM, et al. Observation of amounts of movement practice provided during stroke rehabilitation. Arch Phys Med Rehabil 2009 Oct;90(10):1692-1698 [FREE Full text] [doi: 10.1016/j.apmr.2009.04.005] [Medline: 19801058] 19. Gil-Gómez J, Lloréns R, Alcañiz M, Colomer C. Effectiveness of a Wii balance board-based system (eBaViR) for balance rehabilitation: a pilot randomized clinical trial in patients with acquired brain injury. J Neuroeng Rehabil 2011;8:30 [FREE Full text] [doi: 10.1186/1743-0003-8-30] [Medline: 21600066] https://games.jmir.org/2017/1/e3/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e3 | p. 12 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Prahm et al 20. Herz NB, Mehta SH, Sethi KD, Jackson P, Hall P, Morgan JC. Nintendo Wii rehabilitation (“Wii-hab”) provides benefits in Parkinson's disease. Parkinsonism Relat Disord 2013 Nov;19(11):1039-1042. [doi: 10.1016/j.parkreldis.2013.07.014] [Medline: 23968649] 21. Al-Jumaily A, Olivares R. Electromyogram (EMG) driven system based virtual reality for prosthetic and rehabilitation devices. 2009 Presented at: Proceedings of the 11th International Conference on Information Integration and Web-based Applications & Services; December 14-16, 2009; New York p. 582-586. [doi: 10.1145/1806338.1806448] 22. Oppenheim H, Armiger R, Vogelstein R. WiiEMG: a real-time environment for control of the Wii with surface electromyography. 2010 Presented at: Proceedings of 2010 IEEE International Symposium on Circuits and Systems (ISCAS); May 30-June 2, 2010; Paris, France p. 957-960. [doi: 10.1109/iscas.2010.5537390] 23. de la Rosa R, Alonso A, de la Rosa S, Abasalo D. Myo-Pong: a neuromuscular game for the UVa-Neuromuscular Training System platform. 2008 Presented at: Virtual Rehabilitation ICVR; August 25-27, 2008; Vancouver, BC p. 61-61. [doi: 10.1109/icvr.2008.4625124] 24. Bouwsema H, van der Sluis CK, Bongers RM. The role of order of practice in learning to handle an upper-limb prosthesis. Arch Phys Med Rehabil 2008 Sep;89(9):1759-1764. [doi: 10.1016/j.apmr.2007.12.046] [Medline: 18675393] 25. Sturma A, Herceg M, Bischof B, Fialka-Moser V, Oskar A. Rehabilitation following targeted muscle reinnervation in amputees. In: Replace, Repair, Restore, Relieve--Bridging Clinical and Engineering Solutions in Neurorehabilitation. New York: Springer; 2014:775-779. 26. d'Avella A, Portone A, Fernandez L, Lacquaniti F. Control of fast-reaching movements by muscle synergy combinations. J Neurosci 2006 Jul 26;26(30):7791-7810 [FREE Full text] [doi: 10.1523/JNEUROSCI.0830-06.2006] [Medline: 16870725] 27. Henrichs J, Gagnon M, Hernandez Munoz E, Baker S. SuperTuxKart. 2015. URL: https://supertuxkart.net/Main_Page [accessed 2017-01-23] [WebCite Cache ID 6njJWpdEv] 28. Gramatke S, Gramatke C. Heise. 2015. Pospos - Im Land der Chukchuks URL: https://www.heise.de/download/product/ pospos-im-land-der-chuchuks-74513 [accessed 2017-01-18] [WebCite Cache ID 6nbplEWfs] 29. Danford C, Maynard G. Step Mania 5. 2015. URL: https://www.stepmania.com/download/ [accessed 2017-01-18] [WebCite Cache ID 6nbpCoYec] 30. Ryan RM. Control and information in the intrapersonal sphere: an extension of cognitive evaluation theory. J Pers Soc Psychol 1982;43(3):450-461. [doi: 10.1037//0022-3514.43.3.450] 31. von Held F. Collective Creativity: Exploring Creativity in Social Network Development as Part of Organizational Learning. Wiesbaden: VS Verlag fuer Sozialwissenschaften; 2012. 32. McAuley E, Duncan T, Tammen VV. Psychometric properties of the Intrinsic Motivation Inventory in a competitive sport setting: a confirmatory factor analysis. Res Q Exerc Sport 1989 Mar;60(1):48-58. [doi: 10.1080/02701367.1989.10607413] [Medline: 2489825] 33. Terlaak B, Bouwsema H, van der Sluis CK, Bongers RM. Virtual training of the myosignal. PLoS One 2015 Sep;10(9):e0137161 [FREE Full text] [doi: 10.1371/journal.pone.0137161] [Medline: 26351838] 34. Wurth SM, Hargrove LJ. A real-time comparison between direct control, sequential pattern recognition control and simultaneous pattern recognition control using a Fitts' law style assessment procedure. J Neuroeng Rehabil 2014 May 30;11:91 [FREE Full text] [doi: 10.1186/1743-0003-11-91] [Medline: 24886664] 35. Farina D, Jiang N, Rehbaum H, Holobar A, Graimann B, Dietl H, et al. The extraction of neural information from the surface EMG for the control of upper-limb prostheses: emerging avenues and challenges. IEEE Trans Neural Syst Rehabil Eng 2014 Jul;22(4):797-809. [doi: 10.1109/TNSRE.2014.2305111] [Medline: 24760934] 36. Prahm C, Eckstein K, Ortiz-Catalan M, Dorffner G, Kaniusas E, Aszmann OC. Combining two open source tools for neural computation (BioPatRec and Netlab) improves movement classification for prosthetic control. BMC Res Notes 2016 Aug 31;9(1):429 [FREE Full text] [doi: 10.1186/s13104-016-2232-y] [Medline: 27581624] 37. Jiang N, Dosen S, Muller K, Farina D. Myoelectric control of artificial limbs - is there a need to change focus? [in the spotlight]. IEEE Signal Process Mag 2012 Sep;29(5):152-150. [doi: 10.1109/MSP.2012.2203480] 38. Amsuess S, Vujaklija I, Goebel P, Roche AD, Graimann B, Aszmann OC, et al. Context-dependent upper limb prosthesis control for natural and robust use. IEEE Trans Neural Syst Rehabil Eng 2016 Jul;24(7):744-753. [doi: 10.1109/TNSRE.2015.2454240] [Medline: 26173217] Abbreviations DoF: degree of freedom EMG: electromyographic IMI: Intrinsic Motivation Inventory MVC: maximum voluntary contraction RMS: root mean square https://games.jmir.org/2017/1/e3/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e3 | p. 13 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Prahm et al Edited by G Eysenbach; submitted 25.05.16; peer-reviewed by S Flynn, R Armiger, F Anderson; comments to author 28.06.16; revised version received 18.11.16; accepted 06.01.17; published 09.02.17 Please cite as: Prahm C, Vujaklija I, Kayali F, Purgathofer P, Aszmann OC JMIR Serious Games 2017;5(1):e3 URL: https://games.jmir.org/2017/1/e3/ doi: 10.2196/games.6026 PMID: 28183689 ©Cosima Prahm, Ivan Vujaklija, Fares Kayali, Peter Purgathofer, Oskar C Aszmann. Originally published in JMIR Serious Games (http://games.jmir.org), 09.02.2017. This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Serious Games, is properly cited. The complete bibliographic information, a link to the original publication on http://games.jmir.org, as well as this copyright and license information must be included. https://games.jmir.org/2017/1/e3/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e3 | p. 14 (page number not for citation purposes) XSL FO RenderX http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png JMIR Serious Games JMIR Publications

Game-Based Rehabilitation for Myoelectric Prosthesis Control

Download PDF
14 pages

Loading next page...
 
/lp/jmir-publications/game-based-rehabilitation-for-myoelectric-prosthesis-control-hjQ0zf088j
Publisher
JMIR Publications
Copyright
Copyright © The Author(s). Licensed under Creative Commons Attribution cc-by 4.0
ISSN
2291-9279
DOI
10.2196/games.6026
Publisher site
See Article on Publisher Site

Abstract

Background: A high number of upper extremity myoelectric prosthesis users abandon their devices due to difficulties in prosthesis control and lack of motivation to train in absence of a physiotherapist. Virtual training systems, in the form of video games, provide patients with an entertaining and intuitive method for improved muscle coordination and improved overall control. Complementary to established rehabilitation protocols, it is highly beneficial for this virtual training process to start even before receiving the final prosthesis, and to be continued at home for as long as needed. Objective: The aim of this study is to evaluate (1) the short-term effects of a commercially available electromyographic (EMG) system on controllability after a simple video game-based rehabilitation protocol, and (2) different input methods, control mechanisms, and games. Methods: Eleven able-bodied participants with no prior experience in EMG control took part in this study. Participants were asked to perform a surface EMG test evaluating their provisional maximum muscle contraction, fine accuracy and isolation of electrode activation, and endurance control over at least 300 seconds. These assessments were carried out (1) in a Pregaming session before interacting with three EMG-controlled computer games, (2) in a Postgaming session after playing the games, and (3) in a Follow-Up session two days after the gaming protocol to evaluate short-term retention rate. After each game, participants were given a user evaluation survey for the assessment of the games and their input mechanisms. Participants also received a questionnaire regarding their intrinsic motivation (Intrinsic Motivation Inventory) at the end of the last game. Results: Results showed a significant improvement in fine accuracy electrode activation (P<.01), electrode separation (P=.02), and endurance control (P<.01) from Pregaming EMG assessments to the Follow-Up measurement. The deviation around the EMG goal value diminished and the opposing electrode was activated less frequently. Participants had the most fun playing the games when collecting items and facing challenging game play. Conclusions: Most upper limb amputees use a 2-channel myoelectric prosthesis control. This study demonstrates that this control can be effectively trained by employing a video game-based rehabilitation protocol. (JMIR Serious Games 2017;5(1):e3) doi: 10.2196/games.6026 https://games.jmir.org/2017/1/e3/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e3 | p. 1 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Prahm et al KEYWORDS upper limb prosthesis control; upper extremity amputees; gaming; serious games; neuromuscular rehabilitation; intrinsic motivation; EMG control This game is based on rhythm and speed, and requires a fast Introduction reaction from the player and an immediate transmission of the processed EMG signals to the gaming system. Similarly, a The initial control of a myoelectric prosthesis can be a rehabilitation concept for stroke patients using a modified frustrating experience, especially after the already traumatic version of the WiiMote control is used for rehab purposes of event of losing a limb. Due to the nonintuitive interface, which upper limb amputees, in which EMG signals are matched to the handles a complex mechatronic system, the cognitive demand keys of the WiiMote [22]. However, controls are only limited for controlling the prosthesis is high and further delays the actual to two motions. Other groups chose a game similar to the arcade use of the device in everyday life [1,2]. At least 50% of upper classic Pong, in which the user’s muscle activity is mapped into extremity amputees report problems with prosthesis control and a paddle motion that hits a ball into their opponent’s court [23]. functionality [3,4], which can be attributed to the need for Although those approaches can be motivating, the necessary receiving more training in handling the prosthesis [5,6]. By actions are not very intuitive and are not directly transferable providing more training opportunities, the user can fully benefit to the handling of a prosthesis [24]. from the technical functions of the prosthesis. This study presents an interface between a computer and a To prepare the residual muscles and induce specific brain commercially available surface EMG electrode system plasticity, having access to a prosthesis itself is not necessarily (Ottobock Healthcare GmbH, 13E200), which is commonly needed. Regaining muscle strength and coordination is a used for controlling prostheses, to evaluate the short-term effects cognitively exhausting and repetitive process, during which the on controllability after a video game-based rehabilitation proper execution of movements is reestablished using surface protocol. electromyographic (EMG) feedback [2,7,8]. Through physiotherapy, patients are presented with a variety of tasks This study, compared to previous studies, prompted participants promoting the development of coping strategies for dealing with to not only conduct repetitive agonist and antagonist muscle the activities of daily living, and introducing the embodiment activation, but also to train and exert sustained contractions over of the prosthetic system itself. To effectively control their a short period of time, perform precisely timed contractions, prosthesis, patients need to learn how to properly contract their and elicit simultaneous contraction of both muscles and muscle muscles; the strength of activation and isolation of a single groups. These functions are similar to how patients would muscle are important parameters [1,9,10]. control a real prosthesis as they interact with their environment. Although the standard rehabilitation program offers direct Methods functional benefits, its main shortcomings are the lack of motivation for patients to pursue it without the involvement of Eleven able-bodied participants, who had no prior experience a therapist throughout the lengthy process. In addition to the in EMG control, took part in appraising the benefits of the video loss of functionality, patients may suffer from posttraumatic game-based training. The categories of the EMG controllability depression, further decreasing motivation for rehabilitation [11]. assessments that were evaluated consisted of a provisional Transferring traditional EMG rehabilitation protocols to a virtual maximum voluntary muscle contraction for calibration, precision setting, and incorporating video games into the training process, control, electrode separation, and endurance control by retracing can potentially increase the patient’s engagement and a sine curve with the EMG signal. These assessments are further perseverance [12]. This approach also provides medical explained in detail in the Electromyographic Assessments professionals with quantitative data of the patient’s performance. subsection. Three video games and their respective control Many studies report that the progress achieved during variations were evaluated for their motivational factors and rehabilitation based on a playful concept is faster and superior feasibility. Two questionnaires were given to evaluate (1) the to conservative physiotherapeutic exercises [9,13-15]. These video games and the input method, and (2) intrinsic motivation. rehabilitation games are especially popular in older adults [16], This study was approved by the ethics committee at the Medical and when treating patients affected by stroke [17,18] and University of Vienna (number 1301/2015) and all study Parkinson’s disease [19,20]. Various research groups have participants read and signed the consent form before taking part. addressed adding virtual games to an otherwise dull routine in Participants the area of upper limb amputee rehabilitation. There is, however, Eleven naïve, able-bodied participants without any known a difference between virtual or augmented reality environments neurological or muscular impairments participated in the study. and using commercially available video games during All participants had normal or corrected-to-normal vision and therapeutic interventions [9,21]. The latter provides greater were instructed and accompanied by the examiner throughout accessibility and allows patients to easily set up the games at the entirety of the study. home, and games can be chosen that are proven to motivate the players and maintain engagement over a longer period of time [15]. An example of a commercially available video game that has been interfaced using EMG signals is Guitar Hero [14]. https://games.jmir.org/2017/1/e3/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e3 | p. 2 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Prahm et al Participants were invited two times and had three test sessions Experimental Protocol in total: one Pregaming and Postgaming measurement, both Each participant was seated comfortably in front of two conducted on the same day; and one Follow-Up measurement computer screens. One screen displayed the acquired EMG data to evaluate short-term retention rate two days later. per electrode channel, the other screen showed the game that the participant was playing. Two active surface EMG electrodes Participants were initially instructed to perform three basic EMG (Ottobock Healthcare GmbH 13E200) were positioned on top assessments: the provisional maximum voluntary contraction of the prominent flexor and extensor muscles of the wrist on (MVC) level, accuracy of electrode control, and muscle the participant’s nondominant side (see Figure 1). This was endurance. After a short break, participants were presented with done to match the handedness of the amputees, which is always three computer games in randomized order. After each transferred on the intact limb regardless of the preimpairment EMG-controlled game, they were asked to complete a short state. Amplification and electrode placement remained the same user evaluation survey regarding the gaming experience. After throughout the sessions. Each electrode delivered root mean the third and final game, a modified questionnaire aiming at square (RMS) at the 100-hertz rate of the recorded EMG signal intrinsic motivation (Intrinsic Motivation Inventory; IMI) was following the embedded filtering and rectification. completed. https://games.jmir.org/2017/1/e3/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e3 | p. 3 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Prahm et al Figure 1. The experimental set-up. The outcome measures that were considered included Electromyographic Assessments fluctuations of RMS EMG signals over expected EMG signals. To investigate the changes in overall controllability, three basic Both electrodes typically show some negligible offset activation assessments evaluating approximate strength, muscle precision (due to common noise) during the idle state of the forearm control and coordination, and muscle endurance were performed. muscles (see Figure 2). https://games.jmir.org/2017/1/e3/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e3 | p. 4 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Prahm et al Figure 2. The interface for the precision control assessment and separation of electrode activation. Electrode 1 is used to assess precision control while electrode 2 indicates either the separation or cocontraction of both electrodes. a) The red triangle indicates the goal activation level that participants must reach with their electromyographic signals. b) The black bar marks the threshold at which the electrode is considered active. If the electromyographic signal passed this bar, electrode separation failed and cocontraction is detected. concept of reaching a certain threshold for activating the Maximum Voluntary Contraction Test electrode corresponds to the actual execution of prosthetic The MVC test was used as a calibration of the voltage detected movements [1,25]. The outcome measure was the binary by the electrodes, and assessed the MVC force (averaged over activation of the opposing electrode and the overall percentage 3 trials) for each of the two electrode channels. Participants of activation of the opposing electrode per participant. were asked to maximally contract one muscle and to hold this Assessment of Endurance Control contraction for 1.3 seconds, of which only the last second was taken for calculating the activation baseline. The Assessment of Endurance Control test assessed muscle coordination and muscle fatigue while the participants used Assessment of Precision Control their EMG signals to closely follow a sine curve (1/4 hertz) on The Assessment of Precision Control test evaluated the the screen until they felt fatigued. The estimated force needed participant’s fine EMG control accuracy. The range of this test to reach the peaks of the sine curve corresponds to 60% MVC. was adapted based on the outcome of the MVC test. For each A positive value corresponded to activation of the first electrode, electrode, the participant was asked to reach 30 randomly while a negative value corresponded to the second. Electrode preselected activation levels in the range of 10-90% MVC, and activation needed to be separate to reach the peaks of the sine sustain them for 300 milliseconds each. The required level of curve. The minimum time to be reached in this test was 15 activation was indicated by a triangular mark on the EMG bar minutes. The outcome measure was the EMG signal deviation (see Figure 2 a). A total of 30 marks (3 trials consisting of 10 from the desired sine curve, given as correlation r² [26]. levels) were performed for each electrode. The percentile Games deviation from the mark was taken as outcome measure. Randomization of the goal activation marks took place once Three different open-source games were used in this study: a before the beginning of the study and was kept constant between racing game, a dexterity game, and a rhythm-based game. Each all participants. game featured its own respective control method (see Figure 3). The general input mechanism was to substitute keyboard Assessment of Separation events with EMG activation. Participants controlled two The Assessment of Separation test is a subsection of the electrodes, making two concurrent keys possible at a time. Those Assessment of Precision Control, and determined whether motions represented one degree of freedom (DoF). However, participants could separately control one muscle or if the to allow for more than just one DoF steering, participants could opposing electrode was activated (cocontraction) during the perform a cocontraction (a quick simultaneous activation of tasks of the Precision Control assessment. Depending on the both opposing muscles) and switch to a second DoF. For games MVC, a threshold was set that corresponded to the point of that did not offer the option of cocontraction to switch through EMG activation at which an electrode was considered active keys, the dominant limb supported the control through direct (see Figure 2 b). This threshold was set at 15% MVC. The keyboard input. https://games.jmir.org/2017/1/e3/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e3 | p. 5 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Prahm et al Figure 3. An overview of the three games played by the participants, with their respective input methods and muscle contraction types needed to drive the game. EMG: electromyographic. across the screen. Required EMG activations were quick Racing Game contractions, sustained contractions over a certain time period, In the 3-dimensional racing game Super Tux Kart [27] the player and cocontractions to simultaneously activate two buttons. raced against the clock and computer-controlled adversaries. Questionnaires The participant controlled left and right turns solely with EMG signals, whereas accelerating and braking were controlled using Participants were given two questionnaires to complete: (1) a keyboard inputs with their dominant hand. Required EMG modified IMI questionnaire; and (2) a user evaluation survey activations were quick contractions and sustained contractions. about the EMG assessment, the games that were played, and control methods. Dexterity Game Modified Intrinsic Motivation Inventory Questionnaire In the dexterity game Pospos [28] the player had to maneuver through a 2-dimensional labyrinth and collect items. This game A modified 28-item version of the IMI [30-32] consisting of was controlled entirely through the participant’s EMG signals. five subscales was used to evaluate the experience with the Switching between the DoFs was done by cocontraction, which video games that were played. The five subscales formed scores corresponded to controlling the horizontal and vertical axes of for enjoyment, perceived competence, perceived choice, pressure the player. Required EMG activations were quick contractions, felt, and immersion. An additional six questions were added to sustained contractions, and cocontractions to switch between the last subsection to evaluate immersion into the games. The DoFs. questionnaire was adapted to the study by changing the words, “working” and, “doing” to, “playing”. The questionnaire Rhythm Game included statements such as, “I found the games very In the rhythm-based game Step Mania 5 [29] participants were interesting” and, “I felt tense while playing.” The statements prompted to activate 2 different arrow-shaped buttons using were rated on a 7-point Likert rating scale ranging from 1 (no, their EMG signal. The arrows had to be quickly pressed or held not at all) to 7 (yes, definitely). in time matching the rhythm of the note patterns that scrolled https://games.jmir.org/2017/1/e3/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e3 | p. 6 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Prahm et al Improvement was tested for significance with a related samples User Evaluation Survey Wilcoxon signed rank test. The user evaluation survey consisted of (1) rating the games that were played, (2) rating the input and player control methods Assessment of Endurance Control (see Figure 3), (3) rating the EMG assessment, and (4) The conformity of retracing the sine curve as a correlation r² identifying engaging elements within each game. This short was computed for defined time windows consisting of 30 survey about gaming experience was presented after every game seconds. The highest r² value was taken from a period of at least and included questions about the gameplay, fun factor, 200 seconds. A related samples Wilcoxon signed rank test motivation, and input and control methods. (Cronbach alpha=.05) examined the Pregaming measurement correlation with the Follow-Up measurement correlation. Statistical Analyses All analyses were conducted using IBM SPSS 20 and Matlab Questionnaires 2013b. Nonparametric tests were performed on data not meeting Intrinsic Motivation Inventory Questionnaire the requirement for normal distributions. Normal distributions Participants had to rank the 28 statements from 1 to 7, where 1 were assessed via graphical interpretation showing normal Q-Q represented I do not agree and 7 represented I agree. The plots, and with Shapiro-Wilk tests for normality as a numerical statements belonged to one of five categories and the ranking assessment. Significance was set at Cronbach alpha=.05. was averaged. An independent samples Mann Whitney U test Controllability was performed to describe the data. All categories except pressure had a high desirable rank. Maximum Voluntary Contraction This test was used to relatively set the maximum contraction User Evaluation Survey limit for the subsequent EMG tests, and was given as average This survey consisted of ranked statements on a 5-point scale, of the RMS electrode activation. MVC was measured three and multiple choice questions regarding game experience and times: before playing the games, directly after playing the preferences that were evaluated via a frequency analysis. games, and before the Follow-Up measurement. Results Assessment of Precision Control The outcome measure for this test was the percent deviation Controllability from the 30 goal points per electrode channel. The deviation from the goal was calculated in absolute values and set in Maximum Voluntary Contraction relation to the achieved MVC to derive a percent value of The MVC was used only as a calibration for the electrode deviation for each goal point. All 30 data points were treated channel voltage. However, it could be observed that the RMS as if they were performed consecutively. The goal points were values for the MVC test directly after the games showed an divided into three equidistant intensity sections ranging from increase instead of the expected decrease. Moreover, the same 10% MVC to 90% MVC. The Shapiro-Wilk test confirmed a can be observed for the Follow-Up session. normal distribution with P<.001. The mean and standard Assessment of Precision Control deviation were compared for significance for the activation levels (low, middle, high) and for measurement sessions Results showed a significant improvement in fine accuracy and (Pregaming, Postgaming, Follow-Up) with a Bonferroni electrode coordination from the Pregaming measurement to the corrected paired samples t-test. Follow-Up session (P=.001). Percentile deviation from the goal value was high and heterogeneous in the first two measurements Assessment of Separation (Pregaming and Postgaming), but significantly lower and more Threshold crossings were given in percentages for each of the homogeneous in the Follow-Up measurement (P=.001; see 3 goal activation levels over the 3 measurement sessions. The Figure 4). In all 3 measurement sessions, it was significantly three equidistant intensity levels ranged from 10-90%. harder to reach a high goal activation level compared to a low one (P=.002). https://games.jmir.org/2017/1/e3/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e3 | p. 7 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Prahm et al Figure 4. Development of the deviation of the electrode activation around the goal value levels (separated into low, middle, and high goal values) through all three measurement sessions (Pregaming, Postgaming and Follow-Up). MVC: maximum voluntary contraction. opposing electrode activation from the first to the last Assessment of Separation measurement session for low (P=.04) and middle (P=.02) goal Significantly better performance could be observed during low activation levels; however, this was not true for high intensity level electrode activation tasks compared to high goal activation goal activation (see Figure 5 b). There was a significant tasks within Pregaming (P=.02) and the Follow-Up session improvement in electrode separation overall, considering the (P=.04; see Figure 5 a). There was a significant decrease in first and last measurement sessions (P=.02). Figure 5. Opposing electrode activations over three measurement sessions (Pregaming, Postgaming and Follow-Up) and three goal activation areas, which the participants had to reach with their electromyographic signal (low, middle, and high). a) Comparison within the session. b) Comparison between the sessions. https://games.jmir.org/2017/1/e3/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e3 | p. 8 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Prahm et al shifted by half a period of the retraced sine curve. Nevertheless, Assessment of Endurance Control the Related Samples Wilcoxon Signed Rank Test determined Participants showed an improvement in muscle endurance a significant difference (P=.004) in r² performance between the control (see Figure 6). This result was true for all but one Pregaming and Follow-Up measurement sessions. participant, whose EMG activation expressed an offset and was Figure 6. Scores of the endurance assessment and comparison of Pregaming r² value to the Follow-Up value. High r² corresponds to close electromyographic retracing of the given sine curve. Participants enjoyed playing the games and felt immersed when Questionnaires doing so. Participants perceived playing the games as their own choice and felt competent and at ease while doing so. Pressure Modified Intrinsic Motivation Inventory was the only category in which a low rank was desirable. Results obtained from the IMI questionnaire can be viewed as mean and standard deviation of the five categories in Figure 7. https://games.jmir.org/2017/1/e3/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e3 | p. 9 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Prahm et al Figure 7. Means and standard deviations for 5 subscales of the modified intrinsic motivation questionnaire on a scale from 1 (low) to 7 (high). graphics rated last. Although participants claimed to prefer User Evaluation Survey 3-dimensional graphics, this finding did not reflect their rating Participants were asked to answer five questions for each game of what they enjoyed most in the games. The most motivating (Q1-Q5; see Figure 8). According to the user evaluation survey, aspects of games were (1) the gameplay, (2) to see one’s own the favorite game (derived from the score of Q1 and Q2) was high score, and (3) to clear upgrades. the racing game Super Tux Cart followed by the rhythm game Additionally, participants had to rate the EMG assessments after Step Mania 5. According to Q3 and Q4, participants preferred each session. Participants were asked about how important they to control the games with EMG signals only and to perform thought the EMG assessments were, and to rate the fun they different contraction lengths as well as cocontractions. In terms had while doing them. As can be seen in Figure 9, rating of the of motivation (Q5), the Pospos dexterity game ranked far behind importance of the EMG assessment increased until the the racing and rhythm games, which were equally well received. Follow-Up measurement (however, not significantly), while The most important components to ensure continued play and the participants enjoyed them significantly less (P=.002). enjoyment of a game were (listed according to importance): (1) Interestingly, a slight rise in rating the fun factor was observed the EMG control method, (2) the level of difficulty, (3) dynamic after the Follow-Up session. movements, and (4) collecting items. Music, atmosphere, and Figure 8. Mean and standard deviation ratings of the three games played, according to the survey that participants had to fill in after each game. https://games.jmir.org/2017/1/e3/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e3 | p. 10 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Prahm et al Figure 9. Mean and standard deviation for ratings of the electromyographic assessment after each session, ranging from 5 (very important/fun) to 1 (very unimportant/boring). EMG: electromyographic. the listed EMG tests. However, prolonged exposure to such Discussion stimuli would certainly lead to a loss of interest, which is sure to be maintained by the appealing context of a video game [15]. Results from this study demonstrate improvements in fine accuracy electrode activation and electrode separation from Limitations Pregaming EMG assessments to the Follow-Up measurements. The transferability of the obtained results to the amputee Surprisingly, the MVC values used as a baseline calibration population might be questioned, since this study was conducted also showed an increase, instead of the expected decrease, after strictly with healthy participants. However, based on the playing the games. This result could be due to either warmth outcomes reported in other myocontrol-based studies [34,35], or sweat that would influence the electrode resistance. it is reasonable to expect that the patient group would perform Additionally, this result is a strong indicator that the gaming similarly. session was not fatiguing for the participants. Performance during the precision control assessment, however, declined after This study was a short-term intervention, and can be viewed as playing the games. If, based on previous investigations, we a proof of concept. Further research will incorporate a long-term exclude fatigue, it is reasonable to assume that participants evaluation of video game-based interventions, as well as started losing their concentration by the end of the sessions. In additional exploration of advanced control mechanisms, such the Follow-Up measurement, a clear improvement in as those based on machine learning approaches [36-38]. performance was observed, which can be attributed to the Conclusion restoration of full focus combined with the obtained experience Most upper limb amputees use a 2-channel myoelectric from the previous session. prosthesis control. This study demonstrates that this control can Compared to previous studies [22-24,33], participants not only be effectively trained by employing a video game-based conducted repetitive flexor and extensor muscle activation, but rehabilitation protocol. Participants significantly improved their also sustained contractions over varying periods of time, electrode separation and fine muscle control. It could be shown performed precisely timed contractions, and executed that the enjoyment of the games was greater than that of the simultaneous contractions of both muscle groups. These actions EMG assessments, which decreased over time. Additionally, are similar to how patients would control a real prosthesis. engaging elements within each game could be identified. A subsequent study with an amputee population will show if the The motivational aspects of training gamification are clear, and information gained from healthy participants can be transferred are likely the main advantage compared to conventional to patients. The final outcome would be a robust system that techniques. It is reasonable to assume that certain improvement patients can operate outside of a clinical environment. of the EMG control could be observed by sole application of https://games.jmir.org/2017/1/e3/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e3 | p. 11 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Prahm et al Acknowledgments This work was supported by the Christian Doppler Research Foundation of the Austrian Federal Ministry of Science, Research and Economy, and by the European Research Council Advanced Grant DEMOVE (contract #267888). The authors would like to thank Korbinian Eckstein, MSc for system implementations. Authors CP and IV designed and conducted the study, led the analysis, and interpreted the data; all authors contributed to drafting and reviewing the manuscript. Conflicts of Interest None declared. References 1. Roche AD, Rehbaum H, Farina D, Aszmann OC. Prosthetic myoelectric control strategies: a clinical perspective. Curr Surg Rep 2014 Jan 25;2(3). [doi: 10.1007/s40137-013-0044-8] 2. Sturma A, Göbel P, Herceg M, Gee N, Roche A, Fialka-Moser V, et al. Advanced rehabilitation for amputees after selective nerve transfers: EMG-guided training and testing. In: Replace, Repair, Restore, Relieve – Bridging Clinical and Engineering Solutions in Neurorehabilitation. New York: Springer; 2014:169-177. 3. Kyberd PJ, Beard DJ, Davey JJ, Morrison JD. A survey of upper-limb prosthesis users in Oxfordshire. JPO 1998;10(4):84-91. [doi: 10.1097/00008526-199801040-00004] 4. Dudkiewicz I, Gabrielov R, Seiv-Ner I, Zelig G, Heim M. Evaluation of prosthetic usage in upper limb amputees. Disabil Rehabil 2004 Jan 7;26(1):60-63. [doi: 10.1080/09638280410001645094] [Medline: 14660200] 5. Weeks DL, Anderson DI, Wallace SA. The role of variability in practice structure when learning to use an upper-extremity prosthesis. JPO 2003;15(3):84-92. [doi: 10.1097/00008526-200307000-00006] 6. Roche AD, Vujaklija I, Amsüss S, Sturma A, Göbel P, Farina D, et al. A structured rehabilitation protocol for improved multifunctional prosthetic control: a case study. J Vis Exp 2015 Nov 06(105):e52968. [doi: 10.3791/52968] [Medline: 26575620] 7. Smurr LM, Gulick K, Yancosek K, Ganz O. Managing the upper extremity amputee: a protocol for success. J Hand Ther 2008 Apr;21(2):160-75; quiz 176. [doi: 10.1197/j.jht.2007.09.006] [Medline: 18436138] 8. Aszmann OC, Roche AD, Salminger S, Paternostro-Sluga T, Herceg M, Sturma A, et al. Bionic reconstruction to restore hand function after brachial plexus injury: a case series of three patients. The Lancet 2015 May;385(9983):2183-2189. [doi: 10.1016/s0140-6736(14)61776-1] 9. Anderson F, Bischof W. Augmented reality improves myoelectric prosthesis training. Int J Disabil Hum Dev 2014;13(3):349-354. [doi: 10.1515/ijdhd-2014-0327] 10. Dawson MR, Carey JP, Fahimi F. Myoelectric training systems. Expert Rev Med Devices 2011 Sep;8(5):581-589. [doi: 10.1586/erd.11.23] [Medline: 22026623] 11. Kotila M, Numminen H, Waltimo O, Kaste M. Depression after stroke: results of the FINNSTROKE Study. Stroke 1998 Feb;29(2):368-372 [FREE Full text] [Medline: 9472876] 12. Burke JW, McNeill MDJ, Charles DK, Morrow PJ, Crosbie JH, McDonough SM. Optimising engagement for stroke rehabilitation using serious games. Vis Comput 2009 Aug 27;25(12):1085-1099. [doi: 10.1007/s00371-009-0387-4] 13. Tatla SK, Shirzad N, Lohse KR, Virji-Babul N, Hoens AM, Holsti L, et al. Therapists' perceptions of social media and video game technologies in upper limb rehabilitation. JMIR Serious Games 2015;3(1):e2 [FREE Full text] [doi: 10.2196/games.3401] [Medline: 25759148] 14. Armiger R, Vogelstein R. Air-Guitar Hero: a real-time video game interface for training and evaluation of dexterous upper-extremity neuroprosthetic control algorithms. In: Biomedical Circuits and Systems Conference. 2008 Presented at: BioCAS; Nov 2008; IEEE p. 121-124. [doi: 10.1109/biocas.2008.4696889] 15. Lohse K, Shirzad N, Verster A, Hodges N, Van der Loos HF. Video games and rehabilitation: using design principles to enhance engagement in physical therapy. J Neurol Phys Ther 2013 Dec;37(4):166-175. [doi: 10.1097/NPT.0000000000000017] [Medline: 24232363] 16. van Diest M, Stegenga J, Wörtche HJ, Verkerke GJ, Postema K, Lamoth CJ. Exergames for unsupervised balance training at home: a pilot study in healthy older adults. Gait Posture 2016 Feb;44:161-167. [doi: 10.1016/j.gaitpost.2015.11.019] [Medline: 27004651] 17. Duncan PW, Horner RD, Reker DM, Samsa GP, Hoenig H, Hamilton B, et al. Adherence to postacute rehabilitation guidelines is associated with functional recovery in stroke. Stroke 2002 Jan;33(1):167-177 [FREE Full text] [Medline: 11779907] 18. Lang CE, Macdonald JR, Reisman DS, Boyd L, Jacobson KT, Schindler-Ivens SM, et al. Observation of amounts of movement practice provided during stroke rehabilitation. Arch Phys Med Rehabil 2009 Oct;90(10):1692-1698 [FREE Full text] [doi: 10.1016/j.apmr.2009.04.005] [Medline: 19801058] 19. Gil-Gómez J, Lloréns R, Alcañiz M, Colomer C. Effectiveness of a Wii balance board-based system (eBaViR) for balance rehabilitation: a pilot randomized clinical trial in patients with acquired brain injury. J Neuroeng Rehabil 2011;8:30 [FREE Full text] [doi: 10.1186/1743-0003-8-30] [Medline: 21600066] https://games.jmir.org/2017/1/e3/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e3 | p. 12 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Prahm et al 20. Herz NB, Mehta SH, Sethi KD, Jackson P, Hall P, Morgan JC. Nintendo Wii rehabilitation (“Wii-hab”) provides benefits in Parkinson's disease. Parkinsonism Relat Disord 2013 Nov;19(11):1039-1042. [doi: 10.1016/j.parkreldis.2013.07.014] [Medline: 23968649] 21. Al-Jumaily A, Olivares R. Electromyogram (EMG) driven system based virtual reality for prosthetic and rehabilitation devices. 2009 Presented at: Proceedings of the 11th International Conference on Information Integration and Web-based Applications & Services; December 14-16, 2009; New York p. 582-586. [doi: 10.1145/1806338.1806448] 22. Oppenheim H, Armiger R, Vogelstein R. WiiEMG: a real-time environment for control of the Wii with surface electromyography. 2010 Presented at: Proceedings of 2010 IEEE International Symposium on Circuits and Systems (ISCAS); May 30-June 2, 2010; Paris, France p. 957-960. [doi: 10.1109/iscas.2010.5537390] 23. de la Rosa R, Alonso A, de la Rosa S, Abasalo D. Myo-Pong: a neuromuscular game for the UVa-Neuromuscular Training System platform. 2008 Presented at: Virtual Rehabilitation ICVR; August 25-27, 2008; Vancouver, BC p. 61-61. [doi: 10.1109/icvr.2008.4625124] 24. Bouwsema H, van der Sluis CK, Bongers RM. The role of order of practice in learning to handle an upper-limb prosthesis. Arch Phys Med Rehabil 2008 Sep;89(9):1759-1764. [doi: 10.1016/j.apmr.2007.12.046] [Medline: 18675393] 25. Sturma A, Herceg M, Bischof B, Fialka-Moser V, Oskar A. Rehabilitation following targeted muscle reinnervation in amputees. In: Replace, Repair, Restore, Relieve--Bridging Clinical and Engineering Solutions in Neurorehabilitation. New York: Springer; 2014:775-779. 26. d'Avella A, Portone A, Fernandez L, Lacquaniti F. Control of fast-reaching movements by muscle synergy combinations. J Neurosci 2006 Jul 26;26(30):7791-7810 [FREE Full text] [doi: 10.1523/JNEUROSCI.0830-06.2006] [Medline: 16870725] 27. Henrichs J, Gagnon M, Hernandez Munoz E, Baker S. SuperTuxKart. 2015. URL: https://supertuxkart.net/Main_Page [accessed 2017-01-23] [WebCite Cache ID 6njJWpdEv] 28. Gramatke S, Gramatke C. Heise. 2015. Pospos - Im Land der Chukchuks URL: https://www.heise.de/download/product/ pospos-im-land-der-chuchuks-74513 [accessed 2017-01-18] [WebCite Cache ID 6nbplEWfs] 29. Danford C, Maynard G. Step Mania 5. 2015. URL: https://www.stepmania.com/download/ [accessed 2017-01-18] [WebCite Cache ID 6nbpCoYec] 30. Ryan RM. Control and information in the intrapersonal sphere: an extension of cognitive evaluation theory. J Pers Soc Psychol 1982;43(3):450-461. [doi: 10.1037//0022-3514.43.3.450] 31. von Held F. Collective Creativity: Exploring Creativity in Social Network Development as Part of Organizational Learning. Wiesbaden: VS Verlag fuer Sozialwissenschaften; 2012. 32. McAuley E, Duncan T, Tammen VV. Psychometric properties of the Intrinsic Motivation Inventory in a competitive sport setting: a confirmatory factor analysis. Res Q Exerc Sport 1989 Mar;60(1):48-58. [doi: 10.1080/02701367.1989.10607413] [Medline: 2489825] 33. Terlaak B, Bouwsema H, van der Sluis CK, Bongers RM. Virtual training of the myosignal. PLoS One 2015 Sep;10(9):e0137161 [FREE Full text] [doi: 10.1371/journal.pone.0137161] [Medline: 26351838] 34. Wurth SM, Hargrove LJ. A real-time comparison between direct control, sequential pattern recognition control and simultaneous pattern recognition control using a Fitts' law style assessment procedure. J Neuroeng Rehabil 2014 May 30;11:91 [FREE Full text] [doi: 10.1186/1743-0003-11-91] [Medline: 24886664] 35. Farina D, Jiang N, Rehbaum H, Holobar A, Graimann B, Dietl H, et al. The extraction of neural information from the surface EMG for the control of upper-limb prostheses: emerging avenues and challenges. IEEE Trans Neural Syst Rehabil Eng 2014 Jul;22(4):797-809. [doi: 10.1109/TNSRE.2014.2305111] [Medline: 24760934] 36. Prahm C, Eckstein K, Ortiz-Catalan M, Dorffner G, Kaniusas E, Aszmann OC. Combining two open source tools for neural computation (BioPatRec and Netlab) improves movement classification for prosthetic control. BMC Res Notes 2016 Aug 31;9(1):429 [FREE Full text] [doi: 10.1186/s13104-016-2232-y] [Medline: 27581624] 37. Jiang N, Dosen S, Muller K, Farina D. Myoelectric control of artificial limbs - is there a need to change focus? [in the spotlight]. IEEE Signal Process Mag 2012 Sep;29(5):152-150. [doi: 10.1109/MSP.2012.2203480] 38. Amsuess S, Vujaklija I, Goebel P, Roche AD, Graimann B, Aszmann OC, et al. Context-dependent upper limb prosthesis control for natural and robust use. IEEE Trans Neural Syst Rehabil Eng 2016 Jul;24(7):744-753. [doi: 10.1109/TNSRE.2015.2454240] [Medline: 26173217] Abbreviations DoF: degree of freedom EMG: electromyographic IMI: Intrinsic Motivation Inventory MVC: maximum voluntary contraction RMS: root mean square https://games.jmir.org/2017/1/e3/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e3 | p. 13 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Prahm et al Edited by G Eysenbach; submitted 25.05.16; peer-reviewed by S Flynn, R Armiger, F Anderson; comments to author 28.06.16; revised version received 18.11.16; accepted 06.01.17; published 09.02.17 Please cite as: Prahm C, Vujaklija I, Kayali F, Purgathofer P, Aszmann OC JMIR Serious Games 2017;5(1):e3 URL: https://games.jmir.org/2017/1/e3/ doi: 10.2196/games.6026 PMID: 28183689 ©Cosima Prahm, Ivan Vujaklija, Fares Kayali, Peter Purgathofer, Oskar C Aszmann. Originally published in JMIR Serious Games (http://games.jmir.org), 09.02.2017. This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Serious Games, is properly cited. The complete bibliographic information, a link to the original publication on http://games.jmir.org, as well as this copyright and license information must be included. https://games.jmir.org/2017/1/e3/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e3 | p. 14 (page number not for citation purposes) XSL FO RenderX

Journal

JMIR Serious Games – JMIR Publications

Published: Feb 9, 2017

Keywords: upper limb prosthesis control; upper extremity amputees; gaming; serious games; neuromuscular rehabilitation; intrinsic motivation; EMG control

There are no references for this article.