Hemispheric specialization of functions are tuned by conduction velocities of neuronal propagation in large-scale brain networks

Abstract

AbstractWe propose a tuning-by-delay hypothesis to explain the hemispheric specialization of function that has intrigued psychologists, philosophers, neurologists and neuroscientists for decades. In the auditory domain, speech and melody processing are understood to be lateralized in the left and the right hemispheric brain areas, respectively. Thus, analogous to the notion of pleiotropy where one gene can influence two or three unrelated traits, there exists a shared structural connectome encompassing both hemispheres underlying speech and melody processing, which however gets segregated along left vs right hemispheres, respectively. Here, we demonstrate how empirical observations of hemispheric specialization of speech and melody are shaped by the computational time-scales of information integration in cortical networks. First, we demonstrate that context-specific causal outflow of information emerging from primary auditory cortices (PAC) drives the hemispheric specialization of speech and melody processing when human volunteers listened toa cappellasongs placed within a delayed match to sample task while electroencephalogram (EEG) were recorded. Second, together with participant specific whole-brain connectome model guided by diffusion weighted imaging, we predicted individual specific lateralization indices of inflow in cortical sources - after rigorous source time series reconstruction of EEG spectra. High levels of accuracy in the prediction of laterality indices in the extended large-scale auditory related regions can be achieved after optimizing conduction speeds of information propagation over a neuro-oscillatory network - a novel way to interpret about the neural mechanisms. We demonstrate that parametric modulation of conduction speeds that effectively controls the transmission delays - a key metric for understanding information processing and control of any biological network. Thus, the transmission delay in turn, acts as the switch and triggered by the spectro-temporal complexity of the task context to select the geometry of lateralization.