OBJECTIVE: To explore the feasibility and effectiveness of face recognition-driven video game (FR-VG) swallowing training for stroke patients with dysphagia. DESIGN: A single-blind pilot randomized controlled trial. SETTING: A rehabilitation center. PARTICIPANTS: Stroke patients with dysphagia (N=26). INTERVENTIONS: Participants in the intervention group were trained using FR-VG, while the control group used the conventional method (ie, lip exercise, tongue exercise, and lower jaw exercise). The training was conducted 5 days a week over 4 weeks. OUTCOME MEASURES: The swallowing function and swallowing-related quality of life between the 2 groups were observed before and after the intervention. The acceptance of FR-VG in the intervention group and the adherence of the patients in the 2 groups after the intervention were analyzed. RESULTS: A total of 26 stroke patients with dysphagia were included. The results showed that after the intervention, both groups exhibited significant improvements in swallowing function and swallowing-related quality of life compared to the pre-intervention (P<
.05). Patients in the intervention group demonstrated better swallowing function than the control group on the Gugging Swallowing Screen (P=.015) and Functional Oral Intake Scale (P=.004). The intervention group had a high acceptance of the FR-VG training and had significantly better adherence compared to the control group (P=.032). CONCLUSIONS: FR-VG rehabilitation training can help improve swallowing function, swallowing-related quality of life, and training adherence in stroke patients with dysphagia. Patients have a high acceptance of FR-VG rehabilitation training, which can be used as an adjunctive method to conventional rehabilitation.