Reliability and Validity of Emotrics in the Assessment of Facial Palsy

Abstract

The globally accepted evaluation method for facial palsy is the House–Brackmann facial grading system; however, it does not reflect minute changes. Several methods have been attempted, but there is no universally accepted evaluation method that is non-time-consuming and quantitative. Recently, Emotrics, a two-dimensional analysis that incorporates machine-learning techniques, has been used in various clinical fields. However, its reliability and validity have not yet been determined. Therefore, this study aimed to examine and establish the reliability and validity of Emotrics. All patients had previously received speech therapy for facial palsy at our hospital between January and November 2022. In speech therapy at our hospital, Emotrics was routinely used to measure the state of the patient’s facial palsy. A frame was created to standardize and overcome the limitation of the two-dimensional analysis. Interrater, intrarater, and intrasubject reliability were evaluated with intraclass correlation coefficients (ICC) by measuring the indicators that reflect eye and mouth functions. Validity was evaluated using Spearman’s correlation for each Emotrics parameter and the House–Brackmann facial grading system. A total of 23 patients were included in this study. For all parameters, there was significant interrater and intrarater reliability (ICC, 0.61 to 0.99). Intrasubject reliability showed significant reliability in most parameters (ICC, 0.68 to 0.88). Validity showed a significant correlation in two parameters (p-value < 0.001). This single-center study suggests that Emotrics could be a quantitative and efficient facial-palsy evaluation method with good reliability. Therefore, Emotrics is expected to play a key role in assessing facial palsy and in monitoring treatment effects more accurately and precisely.

Keywords

• facial palsy
• automated evaluation
• emotrics
• interrater reliability
• House–Brackmann facial grading system