Impact of Training on Perception in Dissociated Cortical Culture Neural Networks

Abstract

The way we recall events has an impact on how we perceive them. A good example of this is visual illusion. Can memorized stimuli infl uence the acquisition of a variety of stimuli in tiny neural networks? To answer this question, we used dissociated cortical culture (DCC) grown on a multielectrode array. This in vivo-like in vitro system allows tracking and assessing structural and functional refi nement, as well as the ability for information acquisition, processing, and coding in neural networks. A range of electric stimuli were used to simulate sensory inputs and to train DCC neural circuits. Single, paired-pulse (PP, 20ms ISI) 300mV stimuli at 5, 10, 20, 50, and 100Hz for 1s were delivered in a random time interval. During the analyses, the pre-training, training and testing phases were compared. The data revealed that registered channels increased activity in response to one of the stimulus types while responding less effectively to others. Single, 5Hz, and notably PP stimuli were the favored paradigms. The training phase frequently showed a progressive increase in activity level, with short burst prevalence. However, repetition of the preferred stimuli enhanced the occurrence of both tonic and burst evoked responses with prolonged duration throughout the testing phase. The data demonstrate that even such a tiny network of neurons as we have in DCC is capable for selection of the sensory inputs according to previously memorized information and at their most basic levels they are effi - cient of performing some functions that are typically attributed to higher neural structures.