Taosheng Liu, PhD - US grants

DESCRIPTION (provided by applicant): The environment contains far more information than the brain can process at once. To cope with such information overload, humans need to selectively attend to relevant information and prioritize its processing. In many situations, humans need to select arbitrary features and objects in the scene and maintain attention on the selected information. It is often assumed that an attentional priority signal encodes the current focus of attention and its deployment. However, how the brain computes and maintains attentional priority for features and objects is not known. The long-term goal is to understand how the brain selects different types of information and how selection shapes perception to serve goal-oriented behavior. The objective of this proposal is to delineate the cortical circuitry representing attentional priority for features and objects using functional magnetic resonance imaging (fMRI). Based on recent data obtained in our laboratory, we hypothesize that the dorsal frontoparietal network represents different types of selected information with distinct neural populations, forming a multiplexed representation of attentional priority. In this proposal, we wil test this hypothesis by pursuing four specific aims. First, we will determine the neural representation of attentional priority for visual objects. Second, we will seek to establish a quantitative link between priority signals and task performance. Third, we will determine the relationship between attentional priority signals for features and objects and those for spatial locations. Fourth, we will evaluate the degree of categorical representation of attentional priorit, which is essential for flexible deployment of attention. The proposed research is expected to significantly advance our understanding of how the brain selects and maintains non- spatial information, thus filling in a critical gap in the current scientific knowledge. A deeper understanding of how the brain selects features and objects will provide important constraints for models of attention and can potentially transform our understanding of visual information processing and cognitive control. The proposed research is innovative both in terms of conceptual and methodological advances. Conceptually, the research will test the novel hypothesis that the dorsal frontoparietal network represents attention priority for non-spatial dimensions, challenging the prevailing view that these cortical areas mainly represent spatial information. Methodologically, the application of cutting-edge machine learning and data mining techniques (pattern classification, similarity and clustering analysis) represents a novel approach that more fully exploits the complexity and richness of fMRI data than conventional methods. Finally, the proposed research can make connections to other fields such as category learning and decision making, and suggest interesting future directions to examine common neural processes underlying these cognitive functions.

Cognitive neuroscience research seeks to understand the basic brain mechanisms underlying cognitive and behavioral functions. While there are a variety of available research tools, the majority of these techniques are correlational, in that neural activity (or a proxy of neural activity) is measured while human subjects perform a task. Data from these techniques do not allow causal inference, which requires perturbation of the neural system. Transcranial Magnetic Stimulation (TMS) is currently the leading choice for neural perturbation in humans. With Major Research Instrumentation support from the National Science Foundation, Dr. McAuley and colleagues will purchase a TMS system to enhance research and training in cognitive neuroscience at Michigan State University. TMS can be used to produce temporary disruptions in neural activity or to stimulate the cortex in targeted brain regions. Recent developments in this technology allow image-guided TMS delivery, commonly referred to as neuronavigation. This method allows the TMS coil to be precisely positioned over a specified brain structure based on a person's neuroanatomical data obtained using magnetic resonance imaging (MRI) techniques. This capability is important because, although the structure of the brain is roughly similar across people, the exact anatomical location of neural structures can vary considerably. Targets for disruption/stimulation can be identified by selecting and highlighting the desired structure/locations with the brain. Image-guided (neuronavigated) TMS is quickly becoming a widely-used and standard technology in cognitive neuroscience research. The general value of this technology for cognitive neuroscience is that it is a non-invasive tool that can be coupled with functional and structural MRI data to make causal inferences about normal and disordered brain function that are not possible through fMRI/MRI studies alone.

The participating investigators are all active researchers in the cognitive neurosciences. The new instrumentation will transform Michigan State University's capability to conduct cutting-edge cognitive neuroscience studies that help unravel the neural bases of cognitive function in a diverse set of domains, including perception, attention, memory, cognitive control, and decision making. The acquisition will also more broadly contribute to university-wide neuroscience education and training initiatives at the graduate and undergraduate levels. As part of a new and popular undergraduate major in neuroscience, neuronavigated TMS will be included as one of the methods taught in a required core laboratory course that exposes students to a variety of different neuroscience methods. At the graduate level, the new instrumentation will strengthen introductory and advanced neuroimaging courses by providing new education and novel hands-on training and research opportunities. Michigan State University is a also leader in undergraduate research and training opportunities in STEM areas for women and underrepresented minorities, and the training opportunities created by the new instrumentation will interact synergistically with departmental, interdepartmental and university-wide initiatives. Opportunities for summer research experiences for high school students will be created through established contacts with science teachers in area high schools.

Daily life requires us to navigate an environment that contains much more information than the brain can process at once. Using attention, we can process information that is relevant to the task while ignoring information that is irrelevant to the task. One key component of this type of selective attention is based on feature information. For example, when keeping track of our child at the shopping mall we may focus on the color of the child?s clothing. However, in so doing, we are not completely unaware of other aspects of the environment. How does attention to a particular feature impact the processing of the relevant and irrelevant features? Classical studies have supported simple models based on feature similarity, while more recent work has demonstrated more complex relationships among the features. In the present project, the investigator uses a cross-disciplinary approach that integrates methods from psychophysics, computational modeling, and neuroimaging in order to understand the cognitive and neural aspects of an attentional mechanism called ?surround suppression,? in which the processing of features near the attentional target is weakened. A better understanding of this type of attentional mechanism will have implications for many situations in which humans use visual input to guide behavior, such as education, communication, and human factors engineering. The research project will also provide interdisciplinary training opportunities for graduate and undergraduate students in brain and cognitive sciences.

The investigator team will conduct behavioral and neuroimaging experiments to examine the mechanisms underlying the attentional profile. There are three interrelated objectives. First, the investigators will examine how the attentional profile for simple features changes with stimulus context and task demand. To map out the attentional profile, participants are instructed to attend to a feature to perform a target-detection task while the target feature is systematically varied. Second, the investigators will measure the attentional profile for visual objects defined by the conjunction of two features and examine its sensitivity to task demands. In these experiments, the general hypothesis is that the attentional profile can be flexibly adjusted based on stimulus context and task demand, allowing efficient selection of task-relevant information. In the third objective, the investigators will use functional magnetic resonance imaging to measure neural activity while human participants are engaged in the feature-attention task. Model-based neuroimaging analyses will be used to distinguish between two possible neural mechanisms that can cause attention-induced surround suppression.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.