Hand therapist: A rehabilitation approach based on wearable technology and video gaming

Abstract

Hand impairment after stroke is quite debilitating. Present hand rehabilitation approaches, although useful, are still limited as they often require the constant help of a technician or caregiver and also because they are based on repetitive training which may be demo-tivating [1]. More advanced approaches are in development including the use of robotized systems [2]. Nonetheless, these systems are still expensive and require expert support. Here, we show the proof-of-concept of a hand rehabilitation system, dubbed “hand therapist”, inspired in video gaming devices and software which is comprised of the Myo armband, a home-built robotic glove and Unity3D, a video game development engine. With this approach we aim at a solution that combines performance, low-cost and engagement/motivation in hand therapy. The Myo armband is a ~200 euros wearable device comprised of a 9-axis inertial measurement unit (IMU) and 8 electromyography (EMG) sensors [3]. The device thus provides two kinds of data, spatial and gestural data, respectively. Spatial data informs about the orientation and movement of the user's arm whilst gestural data informs what the user is doing with his/her hand in the form of one of several pre-set poses. All data is communicated via Bluetooth with Unity3D. A ~100 euros robotic glove was built based on a sports glove, nylon wiring, two stepper motors for hand opening/closing, and an Arduino Uno board for control and communication with Unity3D. Touch buttons were also used to detect a fully opened/closed hand. A game was built in Unity 3D for hand training in which the user must grab, hold, transport and drop a cube in several increasingly difficult puzzle levels. In the game the user sees virtual hands/arms that replicates the user's movements such that the user can feel more immersed in the game. Presently, the game works as follows: the user wears the Myo armband in the healthy forearm and executes the required movements, which are translated into the movement of the virtual hand/arm. Hand movements are then replicated into the movement of the robotic glove the user wears in the impaired hand. These movements are also simultaneous translated into the movement of a corresponding virtual hand/arm. So far the system was tested only with healthy subjects, but tests with stroke patients are planned. In the future, we also plan to combine the system with an additional Myo armband unit such that is possible to monitor both hand/arms simultaneously for an improved interaction of the user with the game. In this work we proposed a low-cost (<;500 euros) and engaging solution for hand rehabilitation is possible based on gaming devices and software. Future studies with patients are still required.