事件相关电位早成分及行为学变化在大脑对面孔识别特异性中的作用

Effects of early components of event-related potential and behavior changes on specificity of face recognition

目的:用事件相关电位(event-relatedbrainpotential,ERP)及行为学方法探讨面孔加工是否是特异的,如果是特异的,确定是程度特异性或类别特异性。方法:采用“识别匹配”任务,包括“学习”和“测试”两个阶段,在电脑屏幕上相继呈现人面孔、狗面孔和手机照片(其中一半的照片是倒立的)。受试者在测试阶段的任务是对在学习阶段呈现的刺激尽可能快和准确的按压一个键,未在学习阶段呈现的刺激按压另一个键。各个条件下的行为反应及ERP数据自动保存到电脑中。结果:相对于狗面孔及手机照片,对人面孔反应的识别率最低,反应时间最长,且受刺激倒置的影响最明显。3类刺激均可诱发出N1,P1,P170和N170等ERPs早成分,人面孔诱发的P1和N170潜伏期最长且波幅最大,P170潜伏期最短波幅最大,且倒立人面孔诱发的P1,P170和N170的潜伏期延长,波幅增加最明显。结论:大脑对人面孔的加工具有程度特异性,倒立面孔引起P170和N170潜伏期的延长和波幅的增加可能反应了人面孔识别编码过程中关联特征的破坏。

AIM:To explore whether face recognition is special by using event-related pot ential(ERP),and if special,to ascertain that it is special in degree or in kind. METHODS:Identity matching task was adopted,including two stages of learning an d test.The subject's task in test stage was to press a button on a response-box to any stimulus presented in study stage,and another button to any new stimulus as fast and accurately as possible whether upright or inverted stimuli.ERPs and behavior performances were synchronously recorded in response to the color phot ographs of the three stimuli material including human faces,dogs' faces and hand set pictures in different conditions. RESULTS:Accuracy rate was the lowest and reaction time(RT) was the longest to human faces relative to other objects, and inversion effect for human faces was the biggest.The early components N1,P1,P170 and N170 were affected by stimulus c ategory and orientation presented.The P1 and N170 of human faces had the longest latency and the biggest amplitude, and P170 of human faces had the shortest lat ency and the biggest amplitude,showing hierarchical relations to these component s of the three stimuli.In addition,inverted human face significantly increased P 1,P170 and N170 latencies and most remarkable increase was found in amplitudes. CONCLUSION:Face recognition is special in degree on ERPs and behavior performa nce.Increased amplitudes and/or delayed latencies of P170 and N170 for inversion faces can reflect the disruption of relational features enabling the encoding o f human face identity.