Brain Mechanisms Underlying Behavioral Specificity and Generalization of Short-Term Texture Discrimination Learning.

Zhe Qu,You Wang,Yanfen Zhen,Liping Hu,Yan Song,Yulong Ding
DOI: https://doi.org/10.1016/j.visres.2014.10.017
IF: 1.8
2014-01-01
Vision Research
Abstract:In this study, we used high-density event-related potentials (ERPs) to investigate the brain mechanisms underlying behavioral specificity and generalization of short-term learning of texture discrimination task (TDT). Human adults were trained with TDT for a single session of 1.5 h and their ERPs were measured on the following day. Behavioral performance showed that, after a same amount of exposure of the trained and untrained conditions during EEG session, learning effects were specific to the trained background orientation but generalized across target locations. ERP data, however, revealed both target-location and background-orientation specific changes. While the behavioral background-orientation specificity mainly involved amplitude enhancement of early N2pc over occipital cortex, behavioral target-location generalization was associated with modulation of tempo-spatial configurations of the N2pc component (early-occipital vs. late-parietal/temporal pattern) and decrease of frontal P2 amplitudes for the trained relative to the untrained condition. The earliest visual component C1 did not show specific effects for either background orientation or target location. These results indicated different brain mechanisms underlying the behavioral specificity and generalization of TDT learning. Based on the present findings and literatures, we propose that perceptual learning may induce not only enhancement of relatively early visual selection of the trained target among distractors but also decreases of top-down attention originating from high-level brain center. The reactivation of top-down attention control in some conditions (e.g., the untrained target-location condition) may compensate for the specific effect induced by the early visual selective attention mechanism, leading to generalization or less specificity of perceptual learning in behavioral performance.
What problem does this paper attempt to address?