Abstract. Observations in humans and primates have revealed the
presence of synchronous rhythmic activity between sensorimotor cortex and
spinal cord during maintained voluntary contractions. Similar observations
from studies on the visual system have led to the suggestion that such rhythmicity
and synchronization reflect mechanisms underlying the integration of distributed
neural processes in the nervous system. Therefore, integration of neuronal
activity from distributed sites in different systems may involve common mechanisms
at the level of interacting spiking neurones. At the neuronal level, the
study of neural integration is the study of large scale synaptic integration.
Within single neurones, the spatial-temporal interaction between the timing
of individual inputs and their input location on the dendritic tree results
in the small random fluctuations in membrane potential throughout the cell.
The study of these spatial-temporal interactions is therefore a key aspect
in understanding the computational properties of neurones, i.e. how the large
number of inputs are assimilated into the single output discharge of the
cell. We propose an interdisciplinary research program using simulation studies
combined with multivariate Fourier methods for analysis of stochastic data
to study spatial-temporal integration in single neurones and networks of
neurones. It is hypothesised that weak stochastic temporal correlation amongst
large populations of spike trainsplays an important part in neural integration
within the human central nervous system.