The most common measurement of neuronal responses is the number of spikes emitted. From biophysical experiments it is clear that the time-relations with which spikes are received by a target neuron play a critical role in the likelihood and timing of the target. Over the past 3 decades it has become clear the pattern of spikes over time, whether related to rate variation or to other factors (a debate that is likely to continue for the foreseeable future), carry information. The most obvious aspect of timing in neuronal responses and one that is frequently measured, is latency.
The importance of latency and recruitment order in neuronal recruitment was first considered spinal cord primary motor neurons. For the forebrain, however, observing similar effects is more difficult, so we are left to speculate more. Because rapid medium and long distance communication occurs with spikes, it is important to learn the relations among the various measurements that are used to estimate latency. We recorded neurons in supragranular layers of V1 to study these relations. The supragranular neurons are especially convenient as a testing ground for ideas about latency because the neurons themselves show a wide dynamic range with rapid transitions from nearly absolute silence to firing rates with a large amount of tuning, making them a particularly favorable testing ground for studying how neuronal latency interacts with stimulus features and firing rate. These neurons show very tight tuning in response strength to orientation changes and many other aspects of two-dimensional stimuli. Their latencies vary over a considerable range (~40 ms) as a function of stimulus contrast, and, when the influence of spike count on the measurements of latency are adjusted for the data sampling problems, the latency heavily dependent on stimulus contrast, whereas the spike count is heavily dependent on the spatial distribution of the stimulus. For low firing rates, latency can be misestimated due to the lack of spikes in the sample. This latter problem is can be understood in the light of statistical descriptions of neuronal firing.
BACK
