An important role for high-spatial frequencies in recognition of facial expressions

Reimi Tsurusawa, Yoshinobu Goto, Akihisa Mitsudome, Shozo Tobimatsu

    Research output: Contribution to journalArticlepeer-review

    3 Citations (Scopus)


    We investigated the information processing of facial expression by ERPs to Chernoff's face. Nine healthy, right-handed adult volunteers served as subjects. The Chernoff's face is a simple drawing which is made-up of rich high-spatial frequency components. Neutral and angry faces were determined psychophysically by changing the angles of eyebrow and mouth of Chernoff's face. Three drawings for non-target stimuli (neutral face, angry face, and wheelchair) and a target stimulus (cactus) were used. Stimuli were presented for either 200 or 300 ms in a random order. ERPs were recorded from 10 electrodes according to the international 10-20 system, and were referred to an electrode at the tip of the nose. At least 150 responses were averaged off-line after artifacts rejection. The latency and amplitude of P100 at O1 and O2 were unaffected by the nature of the stimuli. In contrast, the latency of N170 at T5 and T6 for neutral and angry faces was significantly shorter, and its amplitude was larger than those elicited by the object (p<0.05). A slow negative shift was observed over the 230-450 ms time period to the angry face compared with the neutral face. This negative shift was significantly enhanced at a stimulus duration of 300 ms. Our findings suggest that the recognition of facial expression is set between 230 and 450 ms after the appearance of the face and is influenced by the duration of stimulus. Therefore, the high-spatial frequency components of a face appear to be crucial for the recognition of facial expression.

    Original languageEnglish
    Pages (from-to)53-56
    Number of pages4
    JournalInternational Congress Series
    Publication statusPublished - Mar 2005

    All Science Journal Classification (ASJC) codes

    • Medicine(all)


    Dive into the research topics of 'An important role for high-spatial frequencies in recognition of facial expressions'. Together they form a unique fingerprint.

    Cite this