The Laplacian analysis described previously is used to localize cortical activity subserving visual object recognition in humans. In the first of two experiments, subjects are shown pictures of a human face corrupted by varying amounts of noise. After each picture has been presented for 34 ms against a large uniformly illuminated background, the subject is required to report whether or not he saw the face, by pressing a button. The Laplacian response associated with the report that the face is seen differs from that associated with the report that the face is not seen. The difference between these two Laplacian responses has a simple wave form with peak activity at approximately 206 ms after stimulus onset and approximately 196 ms before median reaction time for the button-press report. Its amplitude and polarity, which vary with centre-electrode location over the posterior scalp, are used to construct a map showing the location of cortical activity subserving recognition of the face. This cortical activity localizes to both temporal lobes with some degree of right hemispheric lateralization in right-handed subjects. In the second of the two experiments, subjects are shown the silhouette of a simple shape, such as a triangle, embedded in a large random dot field. Each silhouette is presented for 17 ms. The visibility of the shape is made to vary from trial to trial and the subject is required to report for each trial whether or not the shape is seen. The Laplacian response associated with the report that no shape is seen is very much smaller than that correctly identifying the shape. The difference between these two Laplacian responses has a simple wave form with peak activity at approximately 207 ms after stimulus onset. Its wave form is essentially the same as that associated with face recognition. Its amplitude and polarity, which vary with centre-electrode location over the posterior scalp, are used to construct a map showing the location of cortical activity subserving recognition of the simple shape. This cortical activity localizes to both temporal lobes with strong right hemispheric lateralization in right-handed subjects. The general topography of this activity is similar to that subserving face recognition. Although generally similar, there are measurable differences between the topographies of right temporal lobe activity associated with face recognition and that associated with simple shape recognition.(ABSTRACT TRUNCATED AT 400 WORDS)
展开▼
