视觉经验与皮层神经元感受野特征

视觉经验在高等动物视皮层神经元感受野特征形成的过程中起着重要作用。正常成年动物皮层细胞的感受野具有明显的方位选择性和双眼视差,而缺乏视觉经验的初生动物和在失视条件下成长的动物,其视皮层中方位选择性细胞数量很少,双眼细胞没有视差调谐。后天的视觉训练能够明显地影响和改变幼年动物视皮层感受野的大小、方位和视差调谐。视觉经验的作用在生后4～7周达到高峰,以后逐渐减弱,对成年动物影响不明显。     

猫外膝体神经元兴奋和抑制过程的相互作用

用神经脉冲自动计数的方法,定量地研究了猫外膝体神经元兴奋与抑制过程在时间和空间上的相互作用。1.时间上的相互作用:在亮度变化瞬间,神经细胞的兴奋或抑制过程增强,表现为放电频率暂时地明显增加或减少。这一作用持续的时间(适应时间),瞬变细胞约0.1秒,持续性小感受野细胞约4～8秒,持续性大感受野细胞约40～80秒。2.空间上的相互作用:改变被照射视网膜面积的大小,同时计算神经元的平均放电频率,可以确定外周抑制区的存在、范围和强度。外周抑制可以同时作用于给-撤细胞的给光反应和撤光反应,也可以选择地只抑制其中某一种反应。纯撤光细胞没有外周抑制区。具有大感受野的外膝体神经元,其感受野的大小与背景光的强度有关。在高亮度情况下,感受野变小,有利于改善分辨率;在低亮度下,感受野扩大,有利于提高灵敏度。     

猫外膝体神经元兴奋和抑制过程的相互作用

用神经脉冲自动计数的方法,定量地研究了猫外膝体神经元兴奋与抑制过程在时间和空间上的相互作用。1.时间上的相互作用:在亮度变化瞬间,神经细胞的兴奋或抑制过程增强,表现为放电频率暂时地明显增加或减少。这一作用持续的时间(适应时间),瞬变细胞约0.1秒,持续性小感受野细胞约4～8秒,持续性大感受野细胞约40～80秒。2.空间上的相互作用:改变被照射视网膜面积的大小,同时计算神经元的平均放电频率,可以确定外周抑制区的存在、范围和强度。外周抑制可以同时作用于给-撤细胞的给光反应和撤光反应,也可以选择地只抑制其中某一种反应。纯撤光细胞没有外周抑制区。具有大感受野的外膝体神经元,其感受野的大小与背景光的强度有关。在高亮度情况下,感受野变小,有利于改善分辨率;在低亮度下,感受野扩大,有利于提高灵敏度。     

外膝状体神经元的亮度反应与感受野ON-OFF反应的关系

外膝体是视觉信息进入新皮层的主要通路,其编码亮度信息的神经机制还不清楚.我们采用随机呈现的连续快速变化(50 Hz)的均匀亮度刺激,显著地提高了猫外膝体神经元对均匀亮度的反应强度,通过反相关算法抽提出神经元的亮度反应函数.约81%的神经元的亮度反应函数为单调性上升或下降,有19%的神经元亮度反应函数为V型.通过分析这些神经元对亮度上升和下降的反应强度与感受野ON和OFF反应强度的关系,表明83%的神经元对亮度的反应模式是由其感受野ON-OFF反应的相对强度决定的,其余17%则与其感受野ON-OFF区的兴奋和抑制的变化相关.这些结果揭示了外膝体神经元编码亮度变化的机制.     

研究: 外膝状体神经元的亮度反应与感受野ON-OFF反应的关系

外膝体是视觉信息进入新皮层的主要通路,其编码亮度信息的神经机制还不清楚.我们采用随机呈现的连续快速变化(50 Hz)的均匀亮度刺激,显著地提高了猫外膝体神经元对均匀亮度的反应强度,通过反相关算法抽提出神经元的亮度反应函数.约81%的神经元的亮度反应函数为单调性上升或下降,有19%的神经元亮度反应函数为V型.通过分析这些神经元对亮度上升和下降的反应强度与感受野ON和OFF反应强度的关系,表明83%的神经元对亮度的反应模式是由其感受野ON-OFF反应的相对强度决定的,其余17%则与其感受野ON-OFF区的兴奋和抑制的变化相关.这些结果揭示了外膝体神经元编码亮度变化的机制.     

内容和运动方向感知计算模型

对视野中的物体及运动方向进行感知是视觉感知的基本问题之一,较高级视皮层从V1区的简单细胞开始分为两个通路：“What通路”和“Where通路”．前者对物体的形状、颜色、纹理等内容感知,后者对空间运动速度和方向等感知．本文利用仿脑视觉信息处理计算结构,研究视觉内容和运动方向上的感知计算模型、计算机理和学习算法．该计算模型是一个三层的神经网络,第一层是视觉信号输入层,用于接收外界图像刺激．第二层是神经信息内部表象层,与第一层的网络联结是通过神经元稀疏表象原理自适应形成神经元的感受野．为此,引入Kullback_Leibler散度描述神经元响应的独立性,极小化该代价函数导出网络联结权值的学习算法．从自然图像块中学习得到图像基函数,这些基函数具有局部性、朝向性和带通滤波性．这些性质与生理实验结果中的V1区简单细胞感受野特征相吻合．将这些基函数作为神经元的感受野,并在第三层对较高级视皮层的内容感知和运动感知神经元进行建模．在理想刺激中加入一定量的噪声后,该模型对内容和运动方向的感知仍有较高的准确率和较好的鲁棒性．最后给出的实验仿真结果说明模型的可行性和学习算法的简单有效性．     

初级视皮层神经元非经典感受野研究的新进展

感觉皮层神经元的非经典感受野(简称"外周")对经典感受野(简称"中心")的调节作用广泛存在于哺乳动物中,被认为是感觉皮层神经元的基本特性.以初级视皮层神经元为例,刺激其外周能有效地调节刺激其中心引起的反应,这种作用主要是抑制性的.理解初级视皮层神经元的外周对中心的调节机制能够深入揭示哺乳动物的感觉皮层神经元信息处理的基本原则.本文综述了引起初级视皮层神经元非经典感受野对经典感受野调节作用的神经环路机制和计算模型研究的进展.     

猫纹状皮层神经元整合野的形态和范围

用正弦调制的移动光栅测量了１２０个猫纹状皮层神经元在长度和宽度方向上的空间整合特性。结果表明：（１）大多数细胞的传统感受野具有长宽相近的结构,然而它整合野多数是宽而短或者窄而长的长条形。（２）整合野大小为传统感受野的２－７倍,平均为３．７倍。简单细胞与复杂细胞整合野的大小没有显著差别。（３）随着感受野的视网膜偏心度的增加,整合野大小有逐渐增大的趋势。（４）感受野靠近视网膜垂直中线的细胞,其整合野可     

猫纹状皮层神经元整合野的形状和范围

用正弦调制的移动光栅测量了１２０个猫纹状皮层神经元在长度和宽度方向上的空间整合特性。结果表明：（１）大多数细胞的传统感受野具有长宽相近的结构,然而它们的整合野多数是宽而短或者窄而长的长条形。（２）整合野大小为传统感受野的２—７倍,平均为３．７倍。简单细胞与复杂细胞整合野的大小没有显著差别。（３）随着感受野的视网膜偏心度的增加,整合野大小有逐渐增大的趋势。（４）感受野靠近视网膜垂直中线的细胞,其整合野可以跨越中线进入同侧视野;在两半视野中整合野的范围没有显著差异,但是同侧视野整合野的作用强度明显弱于对侧视野。以上结果提示,初级视皮层神经元能够对大范围内的图形特征进行整合,这种整合作用来自两侧视野。     

反差极性敏感型运动边缘检测神经元的计算模型

在栖类、在和哺乳类某些运动的视觉系统中发现 对刺激反差极性（变亮或变暗）敏感的运动边缘检测神经元。为揭示这类神经元的信息加工原理,以Ｗｉｍｂｂａｕｅｒ等人提出了捍延ｉｎｓｋｅｒ网络为基础,将不同的时空动态受野线性组合,建立了蟾蜍神经节Ｒ３类神经元和家鸽岛豆（ｎＬＭ）神经元的感受野模型,并模拟了Ｒ３凶对蠕虫样刺激物的“头／尾偏爱”特征,细胞对刺激物的选择性以及ｎＬＭ神经元对运动边缘的反差、方向、边缘     

猫视皮层神经元对光栅反应最优方位在视网膜上分布特性的研究

用金属微电极记录了114个猫皮层17、18区细胞对不同方位光栅图象刺激的反应。细胞最优方位与其感受野中心在视网膜的位置间有系统性关系,即最优方位总是倾向于垂直于各感受野中心与视网膜中心区(area centralis)的连线。这一规律对在18区或17、18区全体记录到的细胞而言,有统计意义。 在17、18区内,仅对于感受野位于视网膜离心度(eccentricity)大于9°视角的细胞、具有较窄感受野(宽度小于2.5°)的细胞以及感受野处于视网膜垂直经线附近的细胞,上述规律才有统计意义,而对感受野离心度小于9°的细胞、感受野宽度大于2.5°的细胞以及感受野在倾斜经线附近的细胞,上述规律不明显。     

The Gaussian derivative model for spatial-temporal vision: I. Cortical model

How do we see the motion of objects as well as their shapes? The Gaussian Derivative (GD) spatial model is extended to time to help answer this question. The GD spatio-temporal model requires only two numbers to describe the complete three-dimensional space-time shapes of individual receptive fields in primate visual cortex. These two numbers are the derivative numbers along the respective spatial and temporal principal axes of a given receptive field. Nine transformation parameters allow for a standard geometric association of these intrinsic axes with the extrinsic environment. The GD spatio-temporal model describes in one framework the following properties of primate simple cell fields: motion properties, number of lobes in space-time, spatial orientation. location, and size. A discrete difference-of-offset-Gaussians (DOOG) model provides a plausible physiological mechanism to form GD-like model fields in both space and time. The GD model hypothesizes that receptive fields at the first stage of processing in the visual cortex approximate 'derivative analyzers' that estimate local spatial and temporal derivatives of the intensity profile in the visual environment. The receptive fields as modeled provide operators that can allow later stages of processing in either a biological or machine vision system to estimate the motion as well as the shapes of objects in the environment.     

Receptive fields in the rat piriform cortex.

Current models of odor discrimination in mammals involve molecular feature detection by a large family of diverse olfactory receptors, refinement of molecular feature extraction through precise projections of olfactory receptor neurons to the olfactory bulb to form an odor-specific spatial map of molecular features across glomerular layer, and synthesis of these features into odor objects within the piriform cortex. This review describes our recent work on odor and spatial receptive fields within the anterior piriform cortex and compares these fields with receptive fields of their primary afferent, olfactory bulb mitral/tufted cells. The results suggest that receptive fields in the piriform cortex are ensemble in nature, highly dynamic, and may contribute to odor discrimination and odor memory.     

Parietal neurons encoding visual space in a head-frame of reference.

Neurons of the visual system are known to have receptive fields organized in retinotopic coordinates. We wanted to test whether visual neurons existed whose receptive fields were organized in spatial coordinates. Extracellular recordings from single cells were carried out in one area of the posterior parietal cortex (area V6) of a behaving macaque monkey. Among a great majority of retinotopically organized visual cells, neurons whose visual receptive field did not shift with gaze were also found. These cells responded to the visual stimulation of the same spatial position independently of the animal's direction of gaze, that is, their receptive field was anchored to an absolute spatial location. We suggest that these neurons directly encode visual space and are involved in programming visually-guided motor actions in space.     

Recurrent inhibition and clustered connectivity as a basis for Gabor-like receptive fields in the visual cortex

A continuous-space model of visual cortex interactions which, starting from elongate Gaussian functions, leads to high-selective Gabor-like receptive fields, is proposed. The model is based on intracortical inhibition mechanisms occurring through medium-range clustered connections. The analysis, carried out under the assumption of a linear superposition of geniculate and intracortical contributions, shows how inhibition enhances both spatial and spatial-frequency selectivity. The effects of inhibition strength and of its spatial organization on the properties of the resulting receptive field are investigated. Specifically, the relationships between receptive fields and inhibition schemes are investigated by deriving analytical expressions for their dependence and through a systematic numerical parametric study. The emergence of periodic receptive fields, like the ones observed in neurophysiological measurements, is also pointed out in relation to the clustered nature of the inhibitory scheme.     

视觉系统初级信息加工的一种数学模型——<Ⅰ>模型的空间性质

本文用广义Gabor(EQ)函数作为视觉系统进行初级信息加工的核函数.所谓广义Gabor函数,就是把原始Gabor函数中的一个独立变量扩展到三个(即空间变量从一维推广到二维,再加时间变量),并相应地引入一些参数.本文中EG取二大类形式:各向同性系统中表达式和各向异性系统表达式.当EG取某一特定形式,而且其中参数取某些特定值时,本模型可分别定性描述电生理实验中发现的几种主要感受野类型(Kuffler,Hubel和Wiesel型等).在空间频率域中引入对数变换假设后,EG经付里叶变换后可很好地合成人眼调制传递函数(MTF)     

Independent component filters of natural images compared with simple cells in primary visual cortex.

Properties of the receptive fields of simple cells in macaque cortex were compared with properties of independent component filters generated by independent component analysis (ICA) on a large set of natural images. Histograms of spatial frequency bandwidth, orientation tuning bandwidth, aspect ratio and length of the receptive fields match well. This indicates that simple cells are well tuned to the expected statistics of natural stimuli. There is no match, however, in calculated and measured distributions for the peak of the spatial frequency response: the filters produced by ICA do not vary their spatial scale as much as simple cells do, but are fixed to scales close to the finest ones allowed by the sampling lattice. Possible ways to resolve this discrepancy are discussed.     

Local and global motion preferences in descending neurons of the fly

For a moving animal, optic flow is an important source of information about its ego-motion. In flies, the processing of optic flow is performed by motion sensitive tangential cells in the lobula plate. Amongst them, cells of the vertical system (VS cells) have receptive fields with similarities to optic flows generated during rotations around different body axes. Their output signals are further processed by pre-motor descending neurons. Here, we investigate the local motion preferences of two descending neurons called descending neurons of the ocellar and vertical system (DNOVS1 and DNOVS2). Using an LED arena subtending 240° × 95° of visual space, we mapped the receptive fields of DNOVS1 and DNOVS2 as well as those of their presynaptic elements, i.e. VS cells 1–10 and V2. The receptive field of DNOVS1 can be predicted in detail from the receptive fields of those VS cells that are most strongly coupled to the cell. The receptive field of DNOVS2 is a combination of V2 and VS cells receptive fields. Predicting the global motion preferences from the receptive field revealed a linear spatial integration in DNOVS1 and a superlinear spatial integration in DNOVS2. In addition, the superlinear integration of V2 output is necessary for DNOVS2 to differentiate between a roll rotation and a lift translation of the fly.     

立体图对匹配基元的心理物理学实验的探讨

Marr和Poggio等人从视觉信息加工理论中提出零交叉(Zero-Crossing)是视觉早期加工过程中首先得到的最基本的信息。这在生理上有二个高斯分布叠加的感受野以及侧抑现象等可以作为依据,但在心理物理上仍缺乏佐证。我们利用不同密度的随机点,加上了不同百分比的反差噪声的立体图对,观察了各种条件下体视对噪声的耐受特性,得出了有利于说明黑白相间的边(零交叉)为匹配基元的结果。本文报导了这些实验结果并进行了讨论。     

Visual analysis in unspecialized receptive fields as an orthogonal series expansion

It was postulated on the basis of results obtained by investigating retinal receptive fields of cats that experimental data reflecting the pattern of sensitivity of the receptive field coincide with certain orthogonal polynomials. These special functions have certain properties which can be regarded as spatial characteristics of the receptive fields whose differential sensitivity they describe. Within the framework of the suggested hypothesis visual analysis in the retina from the mathematical point of view can be regarded as a process of expansion of the curve of spatial distribution of brightness within the receptive fields in an orthogonal series. Physiologically speaking the retinal receptive fields can be interpreted as filters distinguishing an elementary orthogonal harmonic from the signal. This method of analysis results in maximal economy of the coding of visual information in the retina. The concepts put forward provide a fresh explanation of a number of physiological facts and they provide precise quantitative evaluations of the response of the receptive field to an arbitrary stimulus.