Yan Xiaoqian, Rossion Bruno
Université de Lorraine, CNRS, CRAN, Nancy, France; Institute of Research in Psychological Science, Institute of Neuroscience, Université Catholique de Louvain, Louvain-La-Neuve, Belgium.
Université de Lorraine, CNRS, CRAN, Nancy, France; Institute of Research in Psychological Science, Institute of Neuroscience, Université Catholique de Louvain, Louvain-La-Neuve, Belgium; Université de Lorraine, CHRU-Nancy, Service de Neurologie, Nancy, France.
Cortex. 2020 Nov;132:281-295. doi: 10.1016/j.cortex.2020.08.016. Epub 2020 Sep 8.
We present an objective and sensitive approach to measure human familiar face recognition (FFR) across variable facial identities. Twenty-six participants viewed sequences of natural images of different unfamiliar faces presented at a fixed rate of 6 Hz (i.e., 6 faces by second), with variable natural images of different famous face identities appearing periodically every 7th image (i.e., .86 Hz). Participants were unaware of the goal of the study and performed an orthogonal task. Following only seven minutes of visual stimulation, the FFR response was objectively identified in the EEG spectrum at .86 Hz and its harmonics (1.71 Hz, etc.) over bilateral occipito-temporal regions, being significant in every individual participant. When the exact same images appeared upside-down, the FFR response amplitude reduced by more than 80%, and was uncorrelated across individuals to the upright face response. The FFR for upright faces emerges between 160 and 200 msec following the famous face onset over bilateral occipito-temporal region and lasts until about 560 msec. The stimulation paradigm offers an unprecedented way to characterize rapid and automatic human face familiarity recognition across individuals, during development and clinical conditions, also providing original information about the time-course and neural basis of human FFR in temporally constrained stimulation conditions with natural images.
我们提出了一种客观且灵敏的方法,用于测量在不同面部身份情况下的人类熟悉面孔识别(FFR)。26名参与者观看了以6赫兹固定速率呈现的不同陌生面孔的自然图像序列(即每秒6张面孔),每隔7张图像(即0.86赫兹)会周期性出现不同著名面孔身份的自然图像。参与者并不知晓该研究的目的,并执行了一项正交任务。仅经过7分钟的视觉刺激后,就在双侧枕颞区域的脑电图频谱中客观地识别出了0.86赫兹及其谐波(1.71赫兹等)处的FFR反应,且在每个参与者中都很显著。当完全相同的图像倒置出现时,FFR反应幅度降低了80%以上,并且个体间与正立面孔反应不相关。正立面孔的FFR在著名面孔出现在双侧枕颞区域后160至200毫秒之间出现,并持续到约560毫秒。该刺激范式提供了一种前所未有的方法,用于在发育和临床情况下跨个体表征快速且自动的人类面孔熟悉度识别,同时也提供了在自然图像的时间受限刺激条件下关于人类FFR的时间进程和神经基础的原始信息。