Chen Y, Bedell H E, Frishman L J
College of Optometry, University of Houston, TX 77204-6052, USA.
Perception. 1996;25(5):505-22. doi: 10.1068/p250505.
Reported differences in neuronal contrast processing between the parallel magnocellular (M) and parvocellular (P) visual pathways invite the hypothesis that contrast discrimination in the human visual system is more sensitive at low contrasts and less sensitive at high contrasts, for stimuli modulated at high compared with low temporal frequencies. In the present study, an edgeless temporally modulated uniform field was selected as the stimulus for psychophysical contrast discrimination, and contrast-increment thresholds for pedestal contrasts ranging from 5.5% to 78.2% were determined with a temporal two-alternative forced-choice staircase procedure. The increment thresholds for five normal subjects were adequately fit by power functions with exponents that shifted continuously from about 0.5 (square-root-law behavior) to about 1.0 (Weber's-law behavior) as stimulus temporal frequency increased from 1 to 30 Hz. A neural simulation, with the use of published contrast-response functions of magnocellular and parvocellular neurons, adjusted with an estimate of response variance, produced two distinct 'neural increment-threshold functions' that were similar to the psychophysical results obtained at the highest and the lowest temporal frequencies, respectively. A shift from a relatively more noise-limited neural mechanism to one whose response is predominantly determined by gain is suggested to account for the change of the contrast-increment-threshold function with increasing temporal frequency.
据报道，在平行的大细胞（M）和小细胞（P）视觉通路之间，神经元对比度处理存在差异，这引发了一种假说，即对于高时间频率调制的刺激与低时间频率调制的刺激相比，人类视觉系统中的对比度辨别在低对比度时更敏感，而在高对比度时则较不敏感。在本研究中，选择了一个无边缘的时间调制均匀场作为心理物理学对比度辨别的刺激，并使用时间二择一强制选择阶梯程序确定了5.5%至78.2%的背景对比度下的对比度增量阈值。随着刺激时间频率从1赫兹增加到30赫兹，五名正常受试者的增量阈值可以通过幂函数得到充分拟合，其指数从约0.5（平方根定律行为）连续变化到约1.0（韦伯定律行为）。利用已发表的大细胞和小细胞神经元的对比度响应函数进行的神经模拟，并根据响应方差估计进行调整，产生了两个不同的“神经增量阈值函数”，分别与在最高和最低时间频率下获得的心理物理学结果相似。随着时间频率的增加，对比度增量阈值函数的变化被认为是由于从相对更多受噪声限制的神经机制向其响应主要由增益决定的机制转变所致。
