Spatiotemporal adaptation through corticothalamic loops--论文代写范文精选

这一直是一个有趣的问题,丘脑皮层反馈是解剖学上占统治地位的,然而,在他们的反应似乎是模棱两可的。实验和理论分析,提出各种操作的视觉皮层,增加相互之间的信息集选区。下面的essay代写范文进行详述。

Abstract 
The thalamus is the major gate to the cortex and its control over cortical responses is well established. Cortical feedback to the thalamus is, in turn, the anatomically dominant input to relay cells, yet its influence on thalamic processing has been difficult to interpret. For an understanding of complex sensory processing, detailed concepts of the corticothalamic interplay need yet to be established. Drawing on various physiological and anatomical data, we elaborate the novel hypothesis that the visual cortex controls the spatiotemporal structure of cortical receptive fields via feedback to the lateral geniculate nucleus. Furthermore, we present and analyze a model of corticogeniculate loops that implements this control, and exhibit its ability of object segmentation by statistical motion analysis in the visual field. Keywords: Lateral geniculate nucleus, Corticothalamic feedback, Control loop, Object segmentation, Motion analysis.

Introduction 
It has long been an intriguing problem that cortical feedback to the thalamus is the anatomically dominant input to relay cells, yet the modulation it effects in their responses seems to be ambiguous. Experiments and theoretical considerations have suggested a variety of operations of the visual cortex on the lateral geniculate nucleus (LGN), such as attention-related gating of geniculate relay cells (GRCs) (Sherman & Koch, 1986; Koch, 1987), synchronizing firing of GRCs (Sillito et al., 1994; Singer, 1994), increasing mutual information between GRCs’ retinal input and their output (McClurkin et al., 1994), and switching GRCs from a detection to an analyzing mode (Sherman & Guillery, 1996; Godwin et al., 1996). Most of the current speculations, however, fall short of a clear, integrated view of corticothalamic function. Detailed concepts of the relationship between thalamic and cortical operation still remain to be established so as to advance our ideas about complex sensory, and ultimately cognitive, processing. 

In the present article we provide support for a role of corticogeniculate loops in complex visual motion processing. In the A laminae of cat LGN two types of X relay cell have been identified that dramatically differ in their temporal response properties (Mastronarde, 1987a; Humphrey & Weller, 1988a; Saul & Humphrey, 1990). Those that are more delayed in response time and phase have been termed lagged, the others nonlagged cells (with the exception of very few so-called partially lagged neurons). In particular, the peak response of lagged neurons to a moving bar is about 100 ms later than that of nonlagged neurons (Mastronarde, 1987a). Lagged X cells comprise about 40 % of all X relay cells (Mastronarde, 1987a; Humphrey & Weller, 1988b). Physiological (Mastronarde, 1987b), pharmacological (Heggelund & Hartveit, 1990), and structural (Humphrey & Weller, 1988b) evidence suggests that rapid feedforward inhibition via intrageniculate interneurons plays a decisive role in shaping the lagged cells’ response. 

Some authors have additionally related differences in receptor types to the lagged-nonlagged dichotomy (Heggelund & Hartveit, 1990; Hartveit & Heggelund, 1990); see, however, Kwon et al. (1991). Layer 4B in cortical area 17 of the cat is the target of both lagged and nonlagged geniculate X cells (Saul & Humphrey, 1992a; Jagadeesh et al., 1997; Murthy et al., 1998). The spatiotemporal receptive fields (RFs) of its direction-selective simple cells can routinely be interpreted as being composed of subregions that receive geniculate inputs alternating between lagged and nonlagged X type (Saul & Humphrey, 1992a; Saul & Humphrey, 1992b; DeAngelis et al., 1995; Jagadeesh et al., 1997; Murthy et al., 1998). At least for simple cells in layer 4B, this RF structure determines the response to moving visual features (McLean & Palmer, 1989; Reid et al., 1991; DeAngelis et al., 1995; Jagadeesh et al., 1993; Jagadeesh et al., 1997; Murthy et al., 1998), and thus the cell’s tuning for direction and speed1 . 

Certainly, intracortical input to cortical cells contributes to direction- selective responses, given that these inputs anatomically outnumber thalamic inputs (Ahmed et al., 1994). Suggested intracortical effects include sharpening of tuning properties by suppressive interactions (Reid et al., 1991; Hirsch et al., 1998; Crook et al., 1998) and amplification of geniculate inputs by recurrent excitation (Douglas et al., 1995; Suarez et al., 1995). Intracortical circuits can in principle even generate their own direction selectivity by selectively inhibiting responses to nonpreferred motion (Douglas et al., 1995; Suarez et al., 1995; Maex & Orban, 1996). Our modeling is complementary to the latter in that we emphasize the influence of geniculate inputs on cortical RF properties that is suggested by numerous studies (Saul & Humphrey, 1992a; Saul & Humphrey, 1992b; Reid & Alonso, 1995; Alonso et al., 1996; Ferster et al., 1996; Jagadeesh et al., 1997; Murthy et al., 1998; Hirsch et al., 1998), in order to bring out effects that are specific to the geniculate contribution to spatiotemporal tuning. 

Thalamocortical neurons possess a characteristic blend of voltage-gated ion channels that jointly determine the timing and pattern of Na+ spiking in response to a sensory stimulus. Depending on the initial membrane polarization, the GRC response to a visual stimulus is in a range between a tonic and a burst mode (Sherman & Guillery, 1996). At hyperpolarization, a Ca2+ conductance gets de-inactivated and, on activation, promotes burst firing. Although the issue is still controversial, there is evidence that a mixture of burst and tonic spikes is involved in the transmission of visual signals in lightly anesthetized or awake animals (Guido et al., 1992; Guido et al., 1995; Guido & Weyand, 1995; Mukherjee & Kaplan, 1995; Sherman & Guillery, 1996; Reinagel et al., 1999). In nonlagged neurons a burst component is present at resting membrane potentials below roughly -70 mV and constitutes a very early part of a visual response (Lu et al., 1992; Guido et al., 1992; Mukherjee & Kaplan, 1995). 

In lagged neurons bursting seems to be responsible for high-activity transients seen after the offset of feedforward inhibition; in particular, it is contributing substantially to the delayed peak response to a moving bar (Mastronarde, 1987b). Cortical feedback to the A laminae of the LGN, arising mainly from layer 6 of area 17 (Sherman & Guillery, 1996), can locally modulate the response mode, and hence the timing, of GRCs by shifting their membrane potentials on a time scale that is long as compared to retinal inputs. This may occur directly through the action of metabotropic glutamate and NMDA receptors (depolarization) (McCormick & von Krosigk, 1992; Godwin et al., 1996; Sherman & Guillery, 1996) and indirectly via the perigeniculate nucleus (PGN) or geniculate interneurons by activation of GABAB receptors (hyperpolarization) of GRCs (Crunelli & Leresche, 1991; Sherman & Guillery, 1996). Indeed, GRCs in vivo are dynamic and differ individually in their degree of burstiness (Lu et al., 1992; Guido et al., 1992; Mukherjee & Kaplan, 1995). Here we explicate the causal link between the variable response timing of GRCs and variable tuning of cortical simple cells for speed of moving features, thus identifying control of speed tuning as a likely mode of corticothalamic operation. Moreover, we exemplify the computational power of the hypothesized control mechanism in a model of corticogeniculate loops that performs object segmentation based on motion cues.

Model of the primary visual pathway 
For the GRCs we have employed a 12-channel model of the cat relay neuron (Huguenard & McCormick, 1992; McCormick & Huguenard, 1992), adapted to 37 degrees Celsius. It includes a transient and a persistent Na+ current, several voltage-gated K+ currents, a voltage- and Ca2+-gated K+ current, a low- and a high-threshold Ca2+ current, a hyperpolarization-activated mixed cation current, and Na+ and K+ leak conductances. As shown in Fig. 1a, retinal input reaches a GRC directly as excitation, and indirectly via an intrageniculate interneuron as inhibition, thus establishing the typical triadic synaptic circuit found in the glomeruli of X GRCs (Sherman & Guillery, 1996). 

The temporal difference between the two afferent pathways equals the delay of the inhibitory synapse and has been taken to be 1.0 ms (Mastronarde, 1987b). It is known that both NMDA and non-NMDA receptors contribute to retinogeniculate excitation to varying degrees, ranging from almost pure non-NMDA to almost pure NMDA mediated responses in individual GRCs of both lagged and nonlagged varieties (Kwon et al., 1991). At least in lagged cells, however, early responses and, hence, responses to the transient stimuli that will be considered here, seem to depend to a lesser degree on the NMDA receptor type than late responses (Kwon et al., 1991). Since the essential characteristics of lagged and nonlagged responses apparently do not depend on the special properties of NMDA receptors – an assumption confirmed by our results – we chose the postsynaptic conductances in GRCs to be of the non-NMDA type.

51Due网站原创范文除特殊说明外一切图文著作权归51Due所有；未经51Due官方授权谢绝任何用途转载或刊发于媒体。如发生侵犯著作权现象，51Due保留一切法律追诉权。（essay代写)
更多essay代写范文欢迎访问我们主页 www.51due.com 当然有essay代写需求可以和我们24小时在线客服 QQ:800020041 联系交流。-X（essay代写)