Department of Otorhinolaryngology, Jena University Hospital, Friedrich-Schiller-University Jena, Am Klinikum 1, 07747, Jena, Germany.
Facial-Nerve-Center Jena, Jena University Hospital, Jena, Germany.
Sci Rep. 2023 Nov 6;13(1):19214. doi: 10.1038/s41598-023-45779-9.
High-resolution facial surface electromyography (HR-sEMG) is suited to discriminate between different facial movements. Whether HR-sEMG also allows a discrimination among the six basic emotions of facial expression is unclear. 36 healthy participants (53% female, 18-67 years) were included for four sessions. Electromyograms were recorded from both sides of the face using a muscle-position oriented electrode application (Fridlund scheme) and by a landmark-oriented, muscle unrelated symmetrical electrode arrangement (Kuramoto scheme) simultaneously on the face. In each session, participants expressed the six basic emotions in response to standardized facial images expressing the corresponding emotions. This was repeated once on the same day. Both sessions were repeated two weeks later to assess repetition effects. HR-sEMG characteristics showed systematic regional distribution patterns of emotional muscle activation for both schemes with very low interindividual variability. Statistical discrimination between the different HR-sEMG patterns was good for both schemes for most but not all basic emotions (ranging from p > 0.05 to mostly p < 0.001) when using HR-sEMG of the entire face. When using information only from the lower face, the Kuramoto scheme allowed a more reliable discrimination of all six emotions (all p < 0.001). A landmark-oriented HR-sEMG recording allows specific discrimination of facial muscle activity patterns during basic emotional expressions.
高分辨率面部表面肌电图(HR-sEMG)适合区分不同的面部运动。HR-sEMG 是否也可以区分面部表情的六种基本情绪尚不清楚。36 名健康参与者(53%为女性,年龄 18-67 岁)被纳入了四个疗程。使用肌肉定位定向电极应用(弗里德伦方案)和基于标志的、与肌肉无关的对称电极布置(栗本方案)同时在面部的两侧记录肌电图。在每个疗程中,参与者根据表达相应情绪的标准化面部图像表达六种基本情绪。同一天重复一次。两周后重复两次疗程,以评估重复效果。HR-sEMG 特征显示,对于两种方案,情绪肌肉激活的系统区域分布模式具有非常低的个体间可变性。当使用整个面部的 HR-sEMG 时,两种方案对于大多数(但不是所有)基本情绪(范围从 p>0.05 到大多数 p<0.001)的不同 HR-sEMG 模式进行统计区分都很好。当仅使用下面部的信息时,栗本方案允许更可靠地区分所有六种情绪(均 p<0.001)。基于标志的 HR-sEMG 记录允许对面部肌肉在基本情绪表达过程中的活动模式进行特定区分。