Muscle coordination patterns are crucial for movement control. Events like strokes can disrupt these patterns. This disruption has been correlated with movement dexterity, which implies that the recruitment of muscle coordination patterns may be an effective target for rehabilitation therapies to restore dexterous movement. In this project, a specific tool was developed to evaluate this potential therapeutic effect. Specifically, a human-computer interface was developed, in which a subject's muscle activity, obtained through surface electromyography, controls a two-dimensional cursor on a screen but only if said activity includes the recruitment of specific muscle coordination patterns. Through visualization of the cursor's movement on the screen, the tool allows for the subject to receive real-time feedback about their muscle coordination patterns, establishing a "human-in-the-loop" feedback control system. The developed system consists of four interconnected modules through standard communication protocols. Surface electromyography signals are obtained using sensors from Delsys, Inc., which deliver data through eight channels at 2kHz. These samples are collected by a module responsible for serving them to an upstream processing module. This second module is responsible for detecting muscle activation through envelope detection and determination of cursor displacement through a movement speed controller. Finally, two application modules were designed. The first one utilizes the processed data to generate a two-dimensional game with clear objectives for the user and the second one displays a real-time history of muscle activations providing information about the patient's performance. Future work will validate the usage of this tool for rehabilitation purposes.