Bowen R W
Department of Psychology, Loyola University Chicago, IL 60626, USA.
Vision Res. 1995 Sep;35(17):2479-90.
To activate selectively cortical ON and OFF pathways, I measured pattern contrast discrimination functions and manipulated contrast polarity (positive and negative) of base contrast (C) and added contrast (delta C). C was a large, long-duration cosine mask and delta C was a brief, localized, spatially narrow-band "D6" pattern. For same polarity C and delta C, contrast discrimination followed a "dipper" pattern: threshold facilitation at low C and a power relation (exponent < 1.0) at high C. The facilitation is predicted from the low-contrast response of cortical neurons and seems to represent isolation of an ON or OFF pathway. Opposite polarity C and delta C give a monotonic function. delta C increases at low base C and remaining higher than the same-polarity function at higher C values. This represents interaction between ON and OFF pathways. Pathway isolation also occurs: a positive test is detected as a contrast increment if masked by negative contrast and a negative test is detected as a contrast decrement if masked by positive contrast. Quantitative aspects of the data suggest a subtractive interaction at low C values and a divisive interaction between pathways at high C values. Test contrast thresholds upon uniform fields of varying luminance show that both the dipper effect and most of the rise in delta C with C are mediated in pattern-selective pathways rather than at a site of luminance adaptation. The pattern-polarity effects on contrast discrimination rule out the "channel uncertainty" explanation for the facilitation dipper. My results suggest that parallel ON and OFF pathways evolved because stimulus-produced decreases in the response of a single pathway are potentially confounded with the effects of contrast adaptation. Thus transient decreases in response in either pathway are not processed and both decreases and increases in contrast are expressed as response increases in separate pathways.
为了选择性激活皮层的开(ON)通路和关(OFF)通路,我测量了模式对比度辨别函数,并操纵了基础对比度(C)和附加对比度(ΔC)的对比度极性(正和负)。C是一个大的、长时间的余弦掩模,而ΔC是一个短暂的、局部的、空间窄带的“D6”模式。对于相同极性的C和ΔC,对比度辨别遵循“勺状”模式:低C时阈值易化,高C时呈幂函数关系(指数<1.0)。这种易化可从皮层神经元的低对比度反应预测得出,似乎代表了ON或OFF通路的分离。相反极性的C和ΔC给出单调函数。在低基础C时ΔC增加,在较高C值时保持高于相同极性函数。这代表了ON和OFF通路之间的相互作用。通路分离也会发生:如果被负对比度掩盖,正测试被检测为对比度增加;如果被正对比度掩盖,负测试被检测为对比度降低。数据的定量方面表明,在低C值时存在减法相互作用,在高C值时通路之间存在除法相互作用。在不同亮度的均匀场地上测试对比度阈值表明,勺状效应以及ΔC随C的大部分增加是在模式选择性通路中介导的,而不是在亮度适应位点。模式极性对对比度辨别的影响排除了对易化勺状现象的“通道不确定性”解释。我的结果表明,平行的ON和OFF通路之所以进化,是因为刺激引起的单个通路反应降低可能与对比度适应的影响混淆。因此,任一通路中反应的短暂降低都不会被处理,对比度的降低和增加都在单独的通路中表现为反应增加。