第33回日本神経科学大会

会期:2010.9.2-4(発表日9月2日)
会場: 神戸コンベンションセンター

Okazaki, Y., Fujita, I. (2010)Responses of Disparity Selective Neurons in V3/V3A to Anti-correlated Random-Dot Stereograms
岡崎安孝、藤田一郎「V3/V3A野における視差選択性細胞の輝度反転ランダムドットステレオグラムに対する応答」

妙録P1-h22
--------------------------------------------------------------------------------

Binocular disparity is a strong cue for depth perception. Most visual areas in the primate brain process binocular disparity. Recent studies by using functional magnetic resonance imaging have demonstrated that among the visual areas, V3/V3A (in macaque monkeys) and V3A (in human subjects) are especially selective for binocular disparity, suggesting V3/V3A is important for stereoscopic depth perception. However, only few electrophysiological studies have examined the characteristics of binocular disparity tuning in V3/V3A neurons. The aim of this study is to characterize the extent to which disparity selective responses in V3/V3A neurons are relevant to depth perception. To examine this, we used correlated and anti-correlated random-dot sterograms (cRDSs and aRDSs, respectively). aRDSs, in which the luminance contrast is inverted in one eye, do not give rise to perception of a plane-in-depth. Therefore, in this case, perceptually relevant neurons should attenuate their disparity selectivity.
We recorded extracellular activity from 117 visually responsive single neurons in V3/V3A of two awake, fixating monkeys. Of these, Ninety five (81%) showed significant selectivity for binocular disparity in cRDSs (Kruskal-Wallis; P < 0.05). We fitted Gabor functions to the data to calculate the ratio of disparity dependent modulation amplitudes between aRDSs and cRDSs. The median ratio for V3/V3A neurons was 0.29, indicating that the modulation by disparity in aRDSs was smaller than that in cRDSs. We compared our results with the ratios obtained for other visual areas. The ratio for V3/V3A was lower than that for V1 (median, 0.39; Cumming & Parker, 1997) (Mann-Whitney test, P = 0.0005) and for V2 (median, 0.33) (P = 0.02), but similar to that for V4 (median, 0.24; Tanabe et al, 2004) (P = 0.74). The results suggest that V3/V3A is situated at a more advanced stage of stereo processing than V1 and V2, but is comparable to V4.