摘要
以新近提出的同心圆感受野模型为基础,从感受野大范围外周区内各亚区之间的抑制性相互作用入手,对视网膜神经节细胞的各种空间传输特性进行了模拟。通过改变外周亚区间抑制性相互作用的敏感度和有效范围,可逼真地模拟神经节细胞的各种不同的面积反应函数,用该模型来处理不同空间频率的正弦光栅时,它既能很好地传递图像的高频成份,又可十分有效地提升被感受野中心/外周拮抗机制所衰减了的低频信息。此外,由于该模型引进了外周亚区间的抑制性相互作用,也解决了其它模型所无法解释的反应强度和感受野结构随图形背景像素的分布密度而变化的这一自适应现象。
Based on a new model of ganglion Cell in Which the inhibitory interactions among subunits of surrounded area of the receptive field (RF) were taken into account, the spatial transfer propCrties of ICtinal ganglion Cells were simulated. By changing either the sensitivity, or the effective range of the inhibitory intCractions, the various area-summation functions were approximated. Tested with a sinusoidal grating consisted of increasing spatial frequencies, the model transferred not only the high frequenCy components (as it was shown by the DOG model), but also the low spatial frequened which were attenuated in the DOG model by the antagonistic inteMCtion between and surround. Moreover, on the basis of the inhibitory interaCtions, the phenomenon that the size of RF and the strength of responses were dependent on the dispersion (density) of background dots can also be Well explained.
出处
《生物物理学报》
CAS
CSCD
北大核心
1995年第3期395-400,共6页
Acta Biophysica Sinica
基金
国家自然科学基金
国家"攀登计划"基金
中国科学院生物物理研究所视觉信息处理开放实验室基金
国家基金委重点项目基金