2017 — 2018 |
Sprague, Thomas C |
F32Activity Code Description: To provide postdoctoral research training to individuals to broaden their scientific background and extend their potential for research in specified health-related areas. |
Effects of Behavioral Priority On Working Memory Representations
PROJECT SUMMARY While much is known about the maintenance of information in working memory (WM), how we prioritize information in WM, and how prioritization changes representations of uncertainty, remain unexplored. This gap in our knowledge is a critical problem because a host of psychiatric and neurologic disorders stems from a primary WM dysfunction. The long-term goal of this work is to understand how neural architectures place limits on WM representations. The overall objective of the proposed research is to test a probabilistic population coding model for how neural activity patterns carry information about WM uncertainty, and elucidate how a network of interacting brain regions sculpt WM representations to satisfy task demands, which is the next step towards my long-term goal. The central hypothesis of the proposed research is that noisy neural activity patterns encode a probability distribution over feature values, whereby greater width of the distribution reflects greater uncertainty about a remembered stimulus feature, and that executive control processes can dynamically prioritize items in WM by sculpting their neural activity pattern to reduce uncertainty. The rationale for the proposed research is that, as we better understand neural constraints limiting WM behavior, a theoretical framework will emerge that can enable strategies for understanding, diagnosing, and treating cognitive dysfunction. Across two specific aims, this research will test the hypotheses that: (1) prioritized items in spatial WM are represented in neural population activity patterns consistent with decreased uncertainty and (2) topographic maps of retinotopic space in frontal, parietal, and occipital cortex, along with non-topographic executive control regions of frontal cortex, contain graded representations of behavioral priority and WM contents. Pariticipants will remember multiple items in visual spatial WM, each assigned different behavioral priority values. Neural activity patterns measured using functional MRI will be used to infer representations of feature values in WM, as well as representations of their uncertainty using a Bayesian generative model of neural activation. Additionally, causal experiments using TMS stimulation and parallel studies in patients with focal lesions to visual, parietal, and frontal cortical regions with topographic maps of space will be conducted to determine the relative role of each region in representing and prioritizing information in WM and its uncertainty. Overall, the proposed work will produce data necessary for testing theories relating neural activity patterns to representations of WM uncertainty across behavioral priority conditions, and the causal involvement of different regions in prioritized WM behavior. The proposed research will be significant because it will uncover how neural representations of uncertainty guide behavior, and how behavioral priority can optimize neural representations of remembered information.
|
0.954 |