Performance for the three simulated models using two possible neural codes: spike count and spike pattern.
We compared the MI between the stimulus and neural activity in individual Ge neurons as a function of the length of stimulus window, using four neural codes: spike count, time-partitioned code, phase-partitioned code combined with spike count and finally combined phase- and time-partitioned codes.
(Left) Decoding performance across Ge n eurons for the intact model using N = 8 bins for each code: spike count (black curve), time-partitioned (blue curve), and phase-partitioned codes (green curve).
We show that CA1 population activity can be described as an evolving traveling wave that exhibits phase coding, rate coding, spike sequences and that generates an emergent population theta rhythm.
However, a rate coding spike count scheme could only explain discrimination between rough vs smooth textures (von Heimendahl et al., 2007), but could not explain discrimination between finer textures well within the psychophysical discrimination curve of the rat (Arabzadeh et al., 2003, 2005; Diamond et al., 2008a).
Coded spikes were stored on a PC at a rate of 1000Hz using CORTEX, a program for neural data collection and analysis developed at the National Institutes of Health (Bethesda, MD).
Coded spikes were stored on a PC at a rate of 1000 Hz using CORTEX, a program for neural data collection and analysis developed at the NIH (Bethesda, MD, USA).
Thus, in the Time- & Phase-partitioned code spikes carry two distinct tags, the first one referring to the position of the spike inside one of the four subdivisions of the stimulus window, the second indicating the phase of the underlying LFP at the moment of the spike occurrence.
This would speed up computations based on spiking responses and focus studies of coding in spike trains on features that are common across all responses to a given stimulus.
To be effective neuronal codes, these spikes should be propagated to axonal terminals where they activate the synapses and drive postsynaptic neurons.
It is known, for instance, that inhibitory GABAergic currents strongly contribute to the temporal coding and spike timing precision of cortical networks during up states of activity [3], [53], [54].
