Study Of Space-time Coding Features Of The Electrical Activity Of Chick Retinal Ganglion Cells

The first stage of visual information processing occurs in the retina. Visual stimuliexcite the photoreceptors, which convert the visual stimuli into electrical signals.These signals are then propagated through the retinal circuitry to the output neurons(the ganglion cells) and are relayed to the brain in the form of the action potentials. Aprimary goal of retinal research is to determine how visual information is encodedand transmitted in the retinal circuitry. In the present study, we investigated thespatiotemporal characteristics of chicken's retinal ganglion cell response. This thesisconsists three parts: how to detect the action potentials and sort the spikes intoneurons'firing activities during multi-electrode recording; how retinal ganglion cellsrespond to flashes with same contrast under the illumination with different level; howthe non-color opponent ganglion cells carry color information by analyzing the crosscorrelation function between nearby retinal ganglion cells.Usually, the electrophysiological recording with planar MEAs (Multi-electrodearray) exhibits a relatively low signal amplitude and low signal-to-noise ratio. In thisstudy, single unit classification was accomplished by performing a nonlinearalgorithm to the data, and utilized the instantaneous frequency and amplitudeinformation simultaneously for the detection of action potential, which resulted in anenhanced signal-to-noise ratio and thus highlighted the action potential peak. Inextracellular recording using flat-mounted electrodes such as MEA (multi-electrodearray), the electric signals picked up by a single electrode may reflex some combinedactivities generated by several adjacent neurons. To investigate the coding propertiesof retinal ganglion cells, the electrode signals should be sorted into the firings ofrelevant neurons. In the present study, spikes were sorted using principal componentanalysis (PCA), the idea behind which is to find an ordered set of orthogonal basisvectors that reflect the directions in the data of largest variation. On comparing theparameters of the first two principal components, the spike data can be classified intoa few clusters with each corresponding to a certain neuron's activity.To explore neural information coding, one expects to find the relationshipbetween the stimuli input and the output of response. However, the same stimulusdoes not always elicit same firing activities. In the visual system, neuron often adaptsto the changes of stimulus by adjusting its sensitivity to the environment. In thepresent study, the activity changes of chicken retinal ganglion cells in response tolight stimuli with defined contrast were investigated, in the presence of various levelsof sustained background illumination. Following a step increase of light illumination,the firing rate of most retinal ganglion cells increased abruptly, and then decreased toa steady-state level with a much lower firing rate during the sustained application oflight. However, when a test flash was applied, which superimposed the prolongedbackground illumination, an increased firing rate was observed. Moreover, theneuron's firing rate was increased to a greater extent when the intensity of thebackground illumination was higher. This might suggest that the neuron sensitivitycould be modified by the background illumination, although the neuron's firing ratewas reduced during sustained illumination. Multi-electrode array (MEA) recorded the electrical activities from a lot ofneurons, which is an advantage for the analyzing of the relationship between nearbyneurons and how the relationship contributed to the information processing. In thepresent study, the electrical activities of paired retinal ganglion cells, under full fieldlight stimuli with a variety of chromatic configuration, were recorded from a smallfunctioning piece of retina using multi-electrode array (MEA). Neurons that hadincreased firings at light-ON and OFF transients and did not show color-opponentproperties were investigated. Single neuron analysis showed that firing rate of eachindividual neuron was dependent on the intensity of illumination. Multi-unit analysisrevealed that adjacent neurons often fired in synchrony in response to lightstimulation. However, in some cases, the strength of correlation between the pairedneurons was higher when the retina was exposed to red or green light with lowintensity, and the correlation was attenuated when yellow or white light with highintensity was given. This seems to suggest that the ensemble activity ofnon-color-opponent ganglion cells might partly participate in color-information...