Inter-hemispheric asymmetry of echoic memory traces with multiple lifetimes

Earlier studies [1,2] suggested exponentially decaying memory trace with significantly higher lifetimes for the association cortex than the primary cortex, and little or no inter-hemispheric differences. In these studies only the peak MEG signal (where signal to noise ratio is largest) or peak strength of the equivalent current dipole was used to model the decay of the echoic memory and quantify its dependence on the inter-stimulus interval (ISI). A unitary cortical response to a stimulus was implied since the shape of the neuromagnetic waveforms was assumed to be insensitive to the ISI. Under this model, the physiological memory trace once established decays passively and uniformly in each cortical area with time thereafter, justifying the use of separate but latency independent lifetimes for the echoic memory in the primary and association auditory cortices. A more general approach was recently proposed [3], which computes the fit to the exponentially saturating function over the wider latency domain. This allows assessing latency-dependent effects that may involve different neuronal pools with different onset latencies and/or memory trace lifetimes. The present study refines the earlier analysis revealing new insights about the latency and especially interhemispheric dependence. The auditory stimuli sequence was modified so that the average response following each ISI comes out from the same number of single trials and mixes together the different past ISIs histories with equal contributions. Our new protocol provides a more objective evaluation of the evoked responses and better dissociates the echoic memory trace decay from refractory effects, which may easily dominate runs with constant ISI. As in one of our previous work [4], the present study makes use of a virtual sensor computation, which maintains the analysis tractable while saving computational effort. 2 Methods