BACKGROUND: Artificial intelligence (AI) is increasingly applied to assess facial reanimation outcomes. However, its clinical utility and evidence have not been systematically reviewed. OBJECTIVE: This study aims to critically appraise studies evaluating AI-based tools used for outcome assessment in patients undergoing facial reanimation. METHODS: This PROSPERO-registered review involved a search of five major databases. We included adult patients with facial palsy of any etiology undergoing static or dynamic facial reanimation procedures, whereby AI tools (e.g., algorithms, simulations) were used in objective and/or subjective outcomes. A qualitative analysis was conducted according to Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) guidelines. RESULTS: Fourteen studies comprising 560 participants were included. The main tools include eight studies using Emotrics to assess structural and symmetry outcomes, demonstrating significant postoperative improvements in oral commissure excursion and symmetry. Five studies using FaceReader demonstrated its strength in capturing socio-emotional dimensions of reanimation and strong correlations with Emotrics and other clinician-rated tools. Two studies showed the utility of Affdex in quantifying the affective dimension of reanimation outcomes. CONCLUSION: By combining precise symmetry analysis with emotional quantification, AI has strong potential to enhance objective and multidimensional assessments of functional and psychosocial outcomes following facial reanimation. Factors affecting its applicability include large heterogeneity and limited validation data.
See et al. (2026) studied this question.