Application of automated face coding (AFC) in older adults: a pilot study

Abstract

Objectives: The study aimed to assess the prevalence and nature of emotional expressions in care-dependent older adults, using an automated face coding (AFC) software. By examining the seven fundamental emotions, the study sought to understand how these emotions manifest and their potential implications for dental care in this population. Methods: Fifty care-dependent older adults’ (mean-age: 78.90 ± 10.83 years; n = 50, men = 25, women = 25) emotional expressions were analyzed using an AFC software. The study measured the prevalence of the seven fundamental emotions including neutral, happy, sad, angry, surprised, scared, and disgusted. Correlations were explored between these expressions and demographic variables such as sex, age, Mini-Mental State Examination (MMSE) scores, as well as the use of sedation. Descriptive statistics, non-parametric tests and Spearman’s rho correlations were applied for statistical analysis (p < 0.05). Results: Neutral expression was the most common emotion (0.732 ± 0.23), with other emotions largely inactive. A trace of happiness was detected in women (0.110 ± 0.23), though not statistically significant (p = 0.061). Significant correlations were found between happy expressions and left eye opening (p = 0.021), and a negative correlation was observed between mouth opening and sad expressions (p = 0.049). No significant associations were found with age, MMSE scores, or sedation use. Conclusions: This study found that AFC software can detect and quantify emotions from facial expressions of dependent older adults and that they predominantly exhibited neutral expressions, with few signs of other emotions. Future research should explore these influences to inform personalized care approaches.