Air-Guitar Hero: A real-time video game interface for training and evaluation of dexterous upper-extremity neuroprosthetic control algorithms

We developed an interface to the commercial video game Guitar Heroreg III using surface electromyography (EMG) to create a novel training and evaluation device for upperextremity amputees. Rather than pressing the keys with onepsilas fingers as in the normal game, in our modified version a user merely flexes his or her index, middle, or ring finger muscles, and the resulting myoelectric activity is recorded using six or more EMG electrodes placed around the forearm. The acquired data is processed in real-time using pattern recognition algorithms to derive intended motion, and the results are used to control the game. Performance metrics reported by the gamepsilas built-in scoring system are used to evaluate classifier performance. To confirm the functionality of the system, three non-amputee users evaluated the EMG-controlled game (called ldquoAir-Guitar Herordquo) and reported that it was effective, fun, and engaging. Ultimately, we intend to use this system as a performance assay for different types of motor decoding algorithms for dexterous control of upper-extremity neuroprostheses.