Background Implementing exercises in the form of video games, otherwise known as exergaming, has gained recent attention as a way to combat health issues resulting from sedentary lifestyles. However, these exergaming apps have not been developed for exercises that can be performed in wheelchairs, and they tend to rely on whole-body movements. Objective This study aims to develop a mobile phone app that implements electromyography (EMG)-driven exergaming, to test the feasibility of using this app to enable people in wheelchairs to perform exergames independently and flexibly in their own home, and to assess the perceived usefulness and usability of this mobile health system. Methods We developed an Android mobile phone app (Workout on Wheels, WOW-Mobile) that senses upper limb muscle activity (EMG) from wireless body-worn sensors to drive 3 different video games that implement upper limb exercises designed for people in wheelchairs. Cloud server recordings of EMG enabled long-term monitoring and feedback as well as multiplayer gaming. Bench testing of data transmission and power consumption were tested. Pilot testing was conducted on 4 individuals with spinal cord injury. Each had a WOW-Mobile system at home for 8 weeks. We measured the minutes for which the app was used and the exergames were played, and we integrated EMG as a measure of energy expended. We also conducted a perceived usefulness and usability questionnaire. Results Bench test results revealed that the app meets performance specifications to enable real-time gaming, cloud storage of data, and live cloud server transmission for multiplayer gaming. The EMG sampling rate of 64 samples per second, in combination with zero-loss data communication with the cloud server within a 10-m range, provided seamless control over the app exergames and allowed for offline data analysis. Each participant successfully used the WOW-Mobile system at home for 8 weeks, using the app for an average of 146 (range 89-267) minutes per week with the system, actively exergaming for an average of 53% of that time (39%-59%). Energy expenditure, as measured by integrated EMG, was found to be directly proportional to the time spent on the app (Pearson correlation coefficient, r=0.57-0.86, depending on the game). Of the 4 participants, 2 did not exercise regularly before the study; these 2 participants increased from reportedly exercising close to 0 minutes per week to exergaming 58 and 158 minutes on average using the WOW-Mobile fitness system. The perceived usefulness of WOW-Mobile in motivating participants to exercise averaged 4.5 on a 5-point Likert scale and averaged 5 for the 3 participants with thoracic level injuries. The mean overall ease of use score was 4.25 out of 5. Conclusions Mobile app exergames driven by EMG have promising potential for encouraging and facilitating fitness for individuals in wheelchairs who have maintained arm and hand mobility.

BACKGROUND Individuals who use wheelchairs face many barriers to regular exercise. Implementing exercises in the form of video games, otherwise known as exergaming, has gained recent attention as a way to increase physical activity and combat health issues resulting from sedentary lifestyles. However, these exergaming apps have not been developed for exercises that can be performed in wheelchairs, and they tend to rely on whole body movements. OBJECTIVE Our objectives were to: 1) develop a mobile phone application that implements EMG-driven exergaming; 2) test the feasibility of using this app to enable people in wheelchairs to perform exergames, used in circuit resistance training, independently and flexibly in their own home; and 3) assess perceived usefulness and useability of this mHealth system. METHODS We have developed a mobile phone application (WOW-Mobile) which senses upper limb muscle activity (EMG) from wireless body-worn sensors to drive 3 different video games which implement upper limb exercises that were designed for people in wheelchairs. The app was developed in Java on an Android platform with the EMG acquired by Flexdots (Dynofit, Inc.), Bluetooth-enabled sensors. EMG drove the game engines and was recorded during all exergaming sessions on our cloud server, in order to enable long-term monitoring and feedback as well as multiplayer gaming. Benchtesting of data transmission and power consumption were tested with Android Studio’s Profiler. Pilot testing was conducted on four individuals with spinal cord injury. Each had a WOW-Mobile system in their home for 8 weeks. By analyzing the session data acquired in real-time and stored on our cloud server, we measured minutes the app was used; minutes exergames were played; and integrated EMG as a measure of energy expended. We also conducted a perceived usefulness and useability questionnaire. RESULTS Benchtest results reveal that the app meets performance specifications to enable real-time gaming, cloud storage of data, and live cloud server transmission for multiplayer gaming. The EMG sampling rate of 64 samples/s, in combination with zero-loss data communication with the cloud server within a 10m range, provided seamless control over the app exergames and allowed for offline data analysis. The WOW-mobile system was pilot tested by four individuals with spinal cord injury. Each participant successfully utilized the WOW-mobile system at home for 8 weeks. They utilized the app for an average of 146 (89-267) minutes per week with the system, actively exergaming for an average of 53% of that time (39-59%). Energy expenditure, as measured by integrated EMG (iEMG) was found to be directly proportional to time spent on the app (Pearson correlation coefficient, r = 0.57-0.86, depending on the game). Two out of the four participants did not exercise regularly before the study; these two participants increased from reportedly exercising close to 0 minutes per week to exergaming 58 and 158 minutes on average, respectively, using the WOW-Mobile fitness system. The perceived usefulness of WOW-mobile in motivating participants to exercise averaged 4.5 on a 5-point Likert-scale, and averaged 5 for the 3 participants with thoracic-level injuries. The average overall ease of use score was 4.25 out of 5. CONCLUSIONS Mobile app exergames driven by EMG has promising potential for encouraging and facilitating fitness for individuals in wheelchairs who have maintained arm and hand mobility. CLINICALTRIAL