User:Vomackae/sandbox

Coding in the Mammalian Gustatory System - Carleton, Accolla, Simon information relating to temporal coding: mainly in CNS -passive vs. active delivery of tastants: no licking/chewing/swallowing with anesthetized animals, affects time of arrival. -active delivery had much faster responses
 * experimental evidence says that among same taste stimulus category, even if rate info (average spike) was the same, temporal info could be used to distinguish between stimuli.
 * changing patterns of neural activity can affect temporal patterns of neural spikes in a way that helps improve discrimination between tastants.
 * study showed that it is possible to discriminate among tastants/concentrations on a single trial, both rate ad temporal info provide more accurate predictions than only rate. (evidence FOR temporal coding)
 * further questions: What is the possible role of temporal dynamics and synchrony in the encoding and learning about taste stimuli in and between brain areas

--Vomackae (talk) 17:42, 31 March 2012 (UTC)

Spikes: Exploring the Neural Code -'understanding' neural code: spike trains of individual neurons convey information, but how? -each image is encoded in patterns of action potentials made by sensory neurons. -'natural signals are presented to us at random, but these signals have correlations that reflect their origins' : stimuli do not always generate the same spike trains, but certain patterns are MORE LIKELY to be generated by the same stimulus. There is an element of randomness, but there is some correlation there. -Problem: interpreting sensory environment based on spike train of one cell. Very difficult, hard to get the big picture as there is a high degree of variability. -Claims of book: 'In a variety of sensory systems, single neurons produce on the order of one spike per characteristic time of stimulus variations - a sparse temporal representation.' Algorithms exist to decode spike trains. Can go from spike to stimulus OR stimulus to spike. (Bayes' rule)

Temporal Encoding: A Vigorous Definition Theunissen and Miller encoding time window: amount of time in which a spike train is to be associated with a certain stimulus 3 tasks of the coding problem: 1)Determine parameters of signal (Ex: encoding time window) 2) AKA the ENcoding problem; Determine encoding technique (temporal vs. rate vs. combination/other) 3) Find a point of reference from which signals can be decoded

Encoding problem is main focus: Different encoding time windows allow for different levels of dimensionality of stimulus. The longer the window, the more complex/difficult to read the stimulus. Also, some spikes may be due to internal noise of system; e.g. temporal coding is not robust. In addition, postsynaptic neural decoder may not be refined enough to extract all the available information encoded in the spike train. Ex: It may group a distinct train of close together spikes into one big stimulus, and as a result, misinterpret the message.

Determining a neural code: Minimal number of symbols necessary to express all biologically significant information. In order to begin to determine symbols, an encoding time window must be determined (Extreme: use the entire spike chain, other extreme: use ms long 'chunks'). Window can be determined by comparing two differing stimulus-response experiments and determining the mutual information of both the stimuli and the spike train responses. Determining encoding schemes involves determining probability that a certain stimulus will elicit a spike pattern (degrees of correlation between stimulus/response patterns). To date, more success has been had with temporal encoding when testing static stimulus/response patterns; e.g. it is difficult to determine encoding window when the stimulus itself is very short.

Temporal coding definition: Pattern of spikes provides more information than number of spikes.

Temporal coding across ensembles of neurons: Research on mammalian gustatory system has shown that there is an abundance of information present in temporal patterns across populations of neurons, and this information is different that the information determined by rate coding schemes. Groups of neurons may synchronize in response to a stimulus. In studies regarding the front cortical portion of the brain in primates (Abeles), precise patterns with short time scales (a few ms) were found across small populations of neurons which correlated with certain information processing behaviors. However, little information could be determined from the patterns; one possible theory is they represented the higher-order processing taking place in the brain. Experiments done with the olfactory system of rabbits (Freeman 1991) showed distinct patterns which correlated with different subsets of odorants. A similar result was obtained in experiments with the locust olfactory system (Laurent and Davidowitz, 1994)

Confusing: Temporal coding vs. temporal encoding: temporal coding = information present with time of stimulus to time of response, both rate and temporal encoding may be used. temporal encoding refers to the pattern of spikes.

The Enigma of the Brain: Zador and Stevens Coding has evolved from strictly rate/frequency to other possibilities of encoding. One of these is temporal coding. Methods of decoding include holding a spike relative to other spikes produced by the same neuron, or relative to the time of onset of the stimulus. Different information could be encoded in each spike. Using the olfactory system as an example, the intensity of the odorant could be conveyed as one or more spikes, while the quality (noxious or pleasant) could be contained in other spike patterns. Evidence supporting temporal coding: Functions of the brain are more temporally precise than mere rate encoding would allow; example: neurons phase-locking to a stimulus and firing within a window of milliseconds. Additional evidence is the variation of spike patterns when exposed to a static stimulus. Neurons in the monkey inferior temporal cortex respond to a one-sec- ond presentation of a complex pattern with a time-varying spike rate, and the spike pattern of the response depend on the shape. This indicates that additional information besides the mean frequency of spikes may be present within a spike train.