Behavioral learning involves modifications to relevant synapses at the neural level, a process known as synaptic plasticity - the ability of the strength of the connections between synapses to change. These changes at the synaptic level are believed to underlie learning and memory. Perceptual learning (PL) is a type of learning that refers to a long lasting improvement in sensory perception as a result of practice or training. PL has been found for a variety of visual tasks, this type of learning also requires synaptic plasticity. However, PL studies typically require a significant amount of training, and concentrate on isolating a single mechanism, leading to great specificity of learning on the trained task. This focus on this specificity has defined the field, but is not representative of ecological conditions. Combining many features is more representation of nature, and recent research suggests this type of training leads to more broad based perceptual benefits. This dissertation will investigate the links between the rules of plasticity at the cellular level and behavior in both a single mechanism, using exposure-based learning, and by combining several PL approaches into a video game framework.

Using the single PL mechanism, exposure-based learning, does not significantly alter behavior on a contrast discrimination task after limited training. Using an integrative approach that combines many perceptual learning mechanisms, including attention, reinforcement, multisensory stimuli, and multi-stimulus dimensions, broad-based benefits of vision were found in a healthy adult population. These results were extended into a highly specialized population, college baseball players. The improvements transferred not only to laboratory tests of vision, but also to improved offensive performance on the baseball field.

Overall, these results give evidence that rules of synaptic plasticity have efficacy when applied at the behavioral level using PL mechanisms. Work remains on establishing optimal training procedures and relating the training induced benefits to the underlying neural mechanisms. Integrating PL approaches that use longer training paradigms produce robust improvements to vision. These findings provide an exciting potential for PL based video-game training to be used as a diagnosis tool and therapeutic intervention to a variety of visual conditions.