Identification of muscle fatigue by tracking facial expressions.

Marco C Uchida,Marcia Calixto Dos Santos,Vitor Daniel Tessutti,Luciane Portas Capelo,Luís Felipe Milano Teixeira,Paulo Henrique Marchetti,Heloiza Prando Ramos,Hélio José Coelho-Júnior,Renato Carvalho,Reury Frank Pereira Bacurau,Bernadette Ann Murphy

doi:10.1371/journal.pone.0208834

Marco C Uchida, Marcia Calixto Dos Santos + Show 9 more

Open Access

https://doi.org/10.1371/journal.pone.0208834

Copy DOI

Abstract

Resistance training (RT) is performed at distinct levels of intensity from the beginning to the end of exercise sets, increasing the sensation of effort as the exercise progress to more vigorous levels, commonly leading to changes on the facial expression of RT practitioners. The objective of this study is to evaluate changes in facial expressions using the Facial Action Coding System(FACS) and the activation of facial muscles by surface electromyography(sEMG) at two different levels of effort during resistance exercise and to investigate the correlation between facial expression and exercise intensity and fatigue. Eleven healthy male participants [23±6years; 1.77±6 m; 78±10kg] performed a set of arm curl exercise at 50% and 85% 1RM until muscle fatigue. The Surface electromyography (sEMG activity was recorded simultaneously in areas of the epicranius muscle (EM) and zygomatic major muscle (ZM). Facial expression was recorded and blindly scored by five experienced examiners. Scores (0–5) were based on the level of activity of the ZM (lip corner puller—Action Unit 12-FACS) during exercise. Facial expression and sEMG data were obtained during the exercise at the first repetition and at muscle failure. The root mean square (RMS) of the sEMG amplitude of the EM was significantly increased between the first and last repetition (50%1RM:p = 0.002,d = 1.75; and 85%1RM:p = 0.002,d = 1.54). The RMS values for the ZM were significantly increased between the first and last repetition (50%1RM:p<0.001,d = 2.67; 85%1RM:p<0.001,d = 0.50). The RMS values for the ZM were also increased in 85%1RM compared to values obtained from 50%1RM (p = 0.001,d = 1.12) at the first repetition. AU12 scores and RMS values were not statistically different between 85%1RM and 50%1RM at the last repetition. Furthermore, there was a strong correlation (r = 0.61;p = 0.045) between AU12 scores and the sEMG peak for the ZM. In conclusion, changes in facial expression may be directly correlated with different resistance exercise intensities and fatigue.

Highlights

In the literature, prior investigations observed that facial expression demonstrated during a physical task acts as a non-verbal behavior able to influence the judgment of observers regarding an athlete’s physical effort [1]
With respect to the available evidence, most investigations evaluated facial expression via the activity of facial muscles assessed by surface electromyography method, and results have demonstrated that facial sEMG activity increased significantly with increasing workload during aerobic exercise [3], and with power output during an incremental workload cycling test [4]
zygomatic major muscle (ZM), which is responsible for the activity of AU12, was studied based on the hypothesis that resistance exercise performed until muscle failure could elicit changes in facial expression similar to those observed in individuals under pain sensation, whose demonstrate drastic changes on AU12 [12]

Summary

Introduction

Prior investigations observed that facial expression demonstrated during a physical task acts as a non-verbal behavior able to influence the judgment of observers regarding an athlete’s physical effort [1]. De Morree and Marcora [2, 3] suggest that sports spectators intuitively assumed the effort exerted by athletes based on their facial expression In this context, inferences can be made about the potential of facial expression to convey physical effort. Unlike sEMG, facial feature tracking has been considered a valuable method to identify patterns of facial movement relating to emotional states, such as sadness, pain, and effort [7,8, 9] In this context, a recent seminal study demonstrated that head movement and upper, lower and whole face position—evaluated trough the facial feature tracking—increased during a laboratory cycling trial according to the exercise intensity [10]

Objectives

Methods

Results

Conclusion