A study to assess the reproducibility of facial expressions used in the Sunnybrook facial grading scale using 4D motion capture technology

Bakshi, Anant (2024). A study to assess the reproducibility of facial expressions used in the Sunnybrook facial grading scale using 4D motion capture technology. University of Birmingham. M.Sc.

Preview

Bakshi024MScByRes_Redacted.pdf
Text - Redacted Version
Available under License All rights reserved.
Download (3MB) | Preview

Abstract

AIM:

The aim of this study was to assess the reproducibility of facial expressions used in the Sunnybrook scale using 4D motion capture technology. This was determined by how well the participants could perform six facial expressions at a 6-month interval (T\(_{1}\) and T\(_{2}\)).

STUDY DESIGN:

Prospective longitudinal cohort study.

MATERIALS AND METHODS:

Twenty healthy participants (10 male & 10 female) meeting the inclusion criteria, were invited to voluntarily participate in the study. They were imaged performing the six facial expressions stated in the Sunnybrook scale (eyebrow lift, eye closed gentle, maximum smile, snarl, lower teeth show and lip pucker) at T\(_{1}\) and T\(_{2}\) (6 months later). All facial expressions were carried out from repose and the volunteers were imaged using a 4D facial motion capture system on both occasions. Upto 8 landmarks were digitally placed on the images, depending on the facial expression, which were then automatically tracked throughout the recording.

The Euclidian distance of each landmark, between T\(_{1}\) and T\(_{2}\), for each facial expression were analysed to determine the magnitude of displacement of the landmarks and determine their reproducibility over the two time points. In addition, for each facial expression the mean absolute displacement of each landmark, in the x, y and z direction between T\(_{1}\) and T\(_{2}\)was determined. A threshold of 2mm was set as being clinically significant different.

RESULTS:

Based on the Euclidian distances, the results of this study showed four of the six facial expressions i.e., snarl, pucker, maximum smile, and eyebrow lift were reproducible. The other two expressions, eye closed gentle and lower teeth show, had differences in Euclidian distance which were statistically significantly different between the two-time intervals.

On further assessment of the x, y and z directions the eyes closed gentle expression was associated with the least reproducibility error, with all distances being statistically significantly less than 2.0mm (p>0.05). None of the six facial expressions had mean absolute differences statistically significantly greater than 2.0mm. The greatest errors in reproducibility were seen with lip pucker were in some directions the upper 95% confidence interval limit exceeded 3.0mm.

CONCLUSIONS:

Four of the facial expressions showed acceptable reproducibility over a 6-month interval, snarl, eyes closed gentle, maximum smile, and eyebrow lift. Lip pucker and lower teeth show were the least reproducible in the x, y and z directions, with the majority of the landmarks being clinically significantly greater than the 2.0mm threshold. The clinical impact of the study is that any changes in facial movement, because of an intervention, need to be greater than 3.0mm to be a true change and not a result of decreased reproduciblity of the facial expression over time.

Type of Work:

Thesis (Masters by Research > M.Sc.)

Award Type:

Masters by Research > M.Sc.

Supervisor(s):