Temporal Coding

Question 1

Why is it important to extract information about the firing of neurons?

Answer 1

To control behaviour, the central nervous system employs approximately one trillion (10^12) neurons, all connected in networks of unfathomable complexity.

Question 2

What has long been thought to be the only means of information coding in neurons?

Answer 2

For decades, most neurophysiologists have assumed that a neuron’s information content is contained solely in its firing rate, the number of action potentials it sends down its axon in any given period

Question 3

What is alternative to rate encoded information?

Answer 3

An alternative view is that temporal firing patterns contain information.

Question 4

Consider the firing pattern of the neuron in the figure. Three groups of 10 action potentials occurring in a 100-ms period travel down

Answer 4

the axon, each group occurring in a different temporal pattern (three insets).

Question 5

In our model, what would rate coding mean?

Answer 5

According to the rate code hypothesis, the timing of each change in firing rate would indicate when an event occurred, and the strength of the increase might report how strong the stimulus was. In each case, however, the “what” of the stimulus would be the same; any single neuron could code for the presence of only a single stimulus property.

Question 6

How would rate coding allow us to determine the strength of the stimulus?

Answer 6

By averaging the firing rates of a number of neurons responding to the same stimulus property, the nervous system could determine the strength of that stimulus at any point in time.

Question 7

How would rate coding allow us to determine the property of the stimulus?

Answer 7

By considering the activity of many such populations responding to different stimulus properties, the exact nature of a complex stimulus could be deciphered.

Question 8

In contrast to the rate code model, the temporal code hypothesis holds that the firing pattern of an individual neuron

Answer 8

could report different “whats,” even while the average firing rate remained unchanged

Question 9

A single neuron like the one in the figure could report the presence of

Answer 9

three different stimuli with the three different temporal firing patterns shown in the insets.

Question 10

What could a temporal code resolve?

Answer 10

some of the apparent ambiguity of the information provided by neurons of the visual cortex.

Question 11

What is the response of each cortical neuron dependent on?

Answer 11

many different features of a stimulus -for example, its orientation, its length, or its contrast.

Question 12

How does a rate code determine the nature of a stimulus?

Answer 12

A rate code requires that the brain determine the exact nature of the stimulus by comparing the output of many different neurons

Question 13

How would a temporal code determine stimulus nature?

Answer 13

in a temporal code, one neuron could unambiguously code changes in a single feature of the stimulus by emitting one of a large repertoire of temporal output patterns (Gawne et al.,1991)

Question 14

What evidence is there for temporal coding in the frontal cortex?

Answer 14

temporal patterns that are spread across two or three neurons recur with precisely defined interspike intervals and may be associated with specific behavioral events (Vaadia, 1995)

Question 15

What evidence is there for temporal coding in the auditory cortex?

Answer 15

Although the average firing rates of groups of neurons in the auditory cortex encode only the onset and offset of longlasting auditory stimuli, the degree of synchronization among groups of neurons (a temporal code) indicates the duration of the stimulus (deCharms et al., 1998)

Question 16

What evidence is there for temporal coding in the visual system?

Answer 16

In the visual system, synchronization among neurons responding to different image features may indicate that those features belong to the same object (Gray et al.,1989).

Question 17

What is the advantage of synchronous firing?

Answer 17

A group of synchronously firing neurons would likely fire their postsynaptic targets more readily than a group of unsynchronized neurons.

Question 18

What evidence is there for temporal coding in retinal ganglion cells and what advantage does this offer?

Answer 18

The synchronous firing of adjacent retinal ganglion cells defines a region closely approximating the overlap between their receptive fields (Berry et al., 1997).

Thus, the retina may use a temporal code to build a more precise representation of a visual image than could be encoded by the firing rates of individual ganglion cells

Question 19

What evidence is there for temporal coding in the hippocampus?

Answer 19

In hippocampal place cells, the phase of action potentials relative to the hippocampal theta rhythm corresponds to whether the animal is entering or leaving the cell’s place field (O’Keefe and Recce, 1993)

Here the timing reference for a spike in one cell is not another spike in the same cell or in other cells, but the phase of an oscillating field potential

Question 20

How would oscillations in the hippocampus be generated?

Answer 20

Such oscillations may be generated by networks of intereurons and provide an important “context” for the- firing of principal neurons (Buzsaki and Chrobak, 1995).

Question 21

What are the implications of the studies mentioned?

Answer 21

All of these results show the existence of unique temporal spike patterns in the brain