| Research Abstract |
We investigated perceptual integration processes in the human brain by using MEG, fMRI, psychophysics and modeling. Following results were obtained.
1. MEG analyses
(1) The MEG responses correlated with the perception of subjective contours were observed aroud V1 in the latency period of 80-150 ms but the responses correlated with figure-ground segmentation were observed around V3/V3A and had longer peck latencies.
(2) It is well known that the responses of neurons in V1 of the primates and cats may be profoundly modulated by the surround stimuli presented outside their classical receptive field. Such effects, recently denoted by a generic term of 'contextual mokulation'. We found contextual modulation of MEG responses in the human V1/V2.
2. fMRI analyses
(1) We developed an integrated software designed to permit computational neuroimaging through the sopport of surface reconstruction, surface edition, surface measurement, and surface-vased sampling of fMRI data.
(2) Neural correlates of color induction were investigated by fMRI. Prominent fMRL responses correlated with induced color were observed in all retinotopic areas, suggesting that the induction may proceed in successive stages of V1-V8.
3. Psychophysics ＆ computational model
(1) A new method for recovery of heading from motion is developed. The simulations of the algorithm were performed, and it was shown that the method edtimates the direction of heading accurately.
(2) Psychophysical experiments on motion assimilation/contrast showed that there are three types of interactions in human motion processing ; one is a summation process effective within nearby regions, and the other two are facilitative and inhibitory induction processes operating over larger spatial scales.
(3) We developed a computational model of scene segmentation based on symmetry axes, a primitive descripton of shape geometry. Human performance and VEP responses were in good agreement with perdictions from the model.
|
