COMPUTATIONAL MODELS OF PFC-MTL INTERACTIONS

点击翻译按钮获取中文摘要

This project will develop a shared computational model to identify unified mechanisms for interactions of the prefrontal cortex (PFC) and medial temporal lobe cortex (MTL) in the performance of a behavioral task requiring cued responses based on context. The specific aims focus on modeling the properties of timing of neural spiking activity in PFC and MTL during performance of related behavioral tasks in rodents in Project 4-5, and on the properties of neural spiking in monkeys in Project 3 and the magnitude of functional magnetic resonance imaging (fMRI) activity in humans in Projects 1-2. In addition, the specific aims will address the effect of the damage to different cortical regions on behavioral performance and measures of neural activity. Based on previous models from this lab, network models of integrate-and-fire neurons or biophysical compartmental simulations representing subregions of the PFC and MTL will be used to generate predictions about experimental data that will guide data analysis, and comparison with the data will determine whether features of the shared model are retained or restructured to account for the data. Modeling predictions will address relative timing of unit activity between regions and timing relative to network oscillations in Projects 4-5, and experimental outcomes will guide selection of neural mechanisms for representation of context, such as oscillatory interference or the temporal context model. Models will generate predictions about the relative timing of unit activity in monkeys in Project 3, and the magnitude of fMRI activity and patterns of preferential viewing in humans in Projects 1 and 2 before and after the transition to context-based responding, and during inferences to cues presented in novel quadrants. The outcomes of these comparisons to data will guide extension of the model to address human and monkey PFC-MTL interactions in relation to rodents.