Correlated Neuronal Activity: High- and Low-Level Views - Computational Neuroscience: A Comprehensive Approach - page 343

to the other randomly, with a characteristic timescale that depends on the experi-
mental setup. The studies in question [36, 37] were done in awake strabismic cats,
a preparation with two advantages: V1 neurons are dominated by a single eye, so
their firing rates essentially depend on what their dominant eye sees regardless of
the other one, and it is relatively easy to know which of the two images is perceived
(at equal contrasts for the two images, one eye always suppresses the other, and
this can be measured by tracking the cat's eye movements in response to conflicting
moving stimuli). The two conditions compared were: a single image presented to
the eye driving the recorded neurons, or the same stimulus shown to the driving eye
plus a conflicting image presented to the other eye. The firing rates in these two
conditions should be the same, because strabismus makes most neurons monocular;
indeed, the rates did not change very much across conditions and did not depend on
which image was perceived. However, synchrony within the 40 Hz band did change
across conditions [36, 37]. When neurons were driven by the eye providing the per-
cept, synchrony was much stronger in the rivalrous condition than in the monocular
one. In contrast, when neurons were driven by the eye whose image became sup-
pressed, synchrony was much lower in the rivalrous condition than in the monocular
one. In other words, when conflicting images were presented, neurons responding
to the image being perceived were always more synchronized. In this case, stronger
synchronization in the high frequency band (30-70 Hz) is suggested to be a neural
correlate of stimulus selection [36, 37].
In summary, it is possible that correlations between neurons can be controlled
independently of firing rate. Two ideas that have been put forth are: that this may
serve to generate more efficient neural codes [71, 44], which follows from theoretical
arguments and experiments in which correlations vary in a stimulus-dependent way,
or to regulate the flow of information [68], which follows from experiments in which
correlations have been linked to expectation, attention, sensory latencies and rivalry
— all processes that regulate the strength but not the content of sensory-derived
neural signals. Other alternatives may become apparent in the future.
Next we discuss some common types of correlated activity patterns. In part, the
goal is to describe them mathematically, at least to a first-order approximation.

12.3 Correlations arising from common input

As mentioned above, oscillations and synchronous responses are commonly ob-
served throughout the nervous system [8-86]. This is not particularly surprising;
in fact, correlations are to be expected simply because neurons in the brain are ex-
tensively interconnected [14, 91]. Now we will discuss two major mechanisms that
give rise to correlated activity, common input and recurent connectivity. The dis-
tinction between them is somewhat artificial, but it is useful in portraying the range
of correlation patterns that may arise. Although they will not be included, it should