First-person perspective gesture observation in virtual reality: A novel approach for anomia rehabilitation in post-stroke aphasia

According to the embodied cognition framework, language is closely linked to the motor system and rehabilitation programs should stimulate language processing through the activation of motor and perceptual systems. In this study, we present an innovative training program based on the embodied cognition framework, using immersive technology to stimulate the perceptual-motor system through gesture observation from a first-person perspective. Post-stroke aphasic patients, in the post-acute phase and presenting with naming deficits, were assigned to one of two training conditions. The experimental group (EG) viewed videos of everyday gestures (i.e., pouring water, cutting carrots) from a first-person perspective, enhancing motor resonance, while the active control group (AC) viewed the same gestures from a third-person perspective. During video playback, the action was named by a female voice and the patient had to repeat it after video completion. The training was administered three times per week for four weeks. Naming abilities, along with other language and quality of life measures, were assessed pre-and post-intervention. Bayesian analysis of the primary outcome-percentage of rehabilitated items-yielded a Bayes factor BF+0 = 2.86, providing anecdotal-to-moderate evidence in favour of the first-person perspective group. The EG showed a higher mean percentage of rehabilitated items (M = 85.5%, SD = 22.5%) compared to the AC group (M = 50.5%, SD = 33.4%). No between-group differences were observed for broader language or quality-of-life measures. These preliminary, hypothesis-generating findings suggest greater efficacy of immersive, embodied gesture observation in promoting word retrieval recovery, and warrant replication in larger, adequately powered studies.