Performance in many perceptual tasks improves with practice. This perceptual learning (PL) has important implications for rehabilitation of patients with visual disorders such as amblyopia, age-related macular degeneration, and focal brain injury, as well as for training of perceptual experts such as radiologists and security screeners. t is also an excellent domain for studying cortical plasticity in general. The improvements are long lasting but often specific to the particular stimuli used in training. This stimulus specificity is universally recognized as a key property of PL, but is still poorly understood. The current PL literature tries to account for all patterns of specificity and transfer exclusively in terms of th amount of overlap in the sensory representations activated and exercised during training and those activated during a subsequent generalization test. Although this idea clearly has merit, such representational overlap is not sufficient for a complete explanation of all cases of transfer Specificity is a major obstacle in rehabilitation and training because the recovered and/or acquired skills may not transfer to new environments. A deeper understanding of the causes of specificity is a necessary precondition for developing methods to overcome it. The perceptual categorization (PC) literature has identified mechanisms that could fill this gap. Converging behavioral, neuropsychological, and neuroimaging evidence indicates that human category learning is mediated by multiple systems. One system is explicit, involves verbal rules, logical reasoning, working memory, and executive attention, and is impaired in frontal patients. Another system is implicit, learns stimulus- response associations, and is impaired in patients with damage to the basal ganglia. The implicit system in COVIS is conceptually similar to the selective-reweighting models that currently dominate the PL literature. The explicit system, however, has heretofore been neglected in PL despite the convergent evidence for its importance in PC. This is problematic because all PL tasks are categorization tasks and typically involve simple unidimensional rules and verbal instructions. The overarching goal of this proposal is to integrate the dual-process approach from PC into the PL literature, with a specific focus on attenuating the negative effects of stimulus specificity of PL. Our working hypothesis is that the implicit system determines the fine-tuned performance after extensive training with a fixed stimulus set, in agreement with current PL models, whereas the explicit system supports much of the generalization to novel stimuli and tasks. We propose the following specific aims.
Aim 1 is to develop a comprehensive model that can account for all patterns of specificity and transfer in PL using the computational cognitive neuroscience framework.
Aim 2 is to collect behavioral data on the role of the explicit and implicit processing systems in PL and across multiple locations and attentional manipulations. The data will be used to test the model and to facilitate further model development.
Perceptual learning (PL) has important implications for rehabilitation of patients with visual disorders (e.g., amblyopia), as well as for training of perceptual experts (e.g., radiologists and security screeners). A key property of PL is stimulus specificity which hinders the development of rehabilitation and training programs. The overall goal of the proposed research is to bring constructs from perceptual category learning into models and experimental studies of perceptual learning with the aims of (a) better understanding the mechanisms associated with stimulus specificity, and (b) developing procedures to attenuate stimulus specificity effects.