This paper addresses the challenges of conventional rehabilitation in stroke patients by proposing a virtual reality rehabilitation system supported by surface electromyography and force measurement. The objective was to select appropriate conventional therapies, adapt them to virtual reality games, and simultaneously acquire electromyography signals from the biceps, triceps, and finger flexors, also we used a force sensor to evaluate patient force. The study aimed to evaluate the platform’s effectiveness and propose a system for clinical monitoring of patient progress. Tests were conducted with two patients using three virtual reality games. The physiological signals were processed by the Teager–Kaiser energy operator and descriptive statistics, this processing of the signals and force information revealed remarkable differences, suggesting a potential method for evaluating and following up on therapy. The study concluded that simultaneous measurement of muscle activity during task performance could be used to monitor progress or evidence of neuronal damage in patients and to suggest new therapies.