To provide a more friendly and interesting rehabilitation experience for wrist rehabilitation, this paper presents an emotional interaction method based on secondary fuzzy reasoning and integrates gamification experience to build emotional interaction hardware equipment for wrist rehabilitation. First, fuzzy inference rule base one is established with multidimensional physiological signals as the input and arousal valence as the output. A multidimensional physiological signal-arousal valence (MPS-AV) protocol is created to calculate the value of arousal valence. Second, fuzzy inference rule base two is established with the arousal valence as the input and the emotional label as the output. An arousal valence–emotion (AV-emotion) protocol is created, and the emotional state of the subjects is further visualized and analyzed. Finally, by encoding the subject’s behavior, the degree of transformation of the subject’s interactive behavior and the success rate of task completion can be calculated to verify the effectiveness of the interaction hardware and emotional interaction method. This provides methods and experience for the design and development of wrist rehabilitation and related digital medical products.