NCS-FO: Uncovering the cognitive and neural fingerprints that make each of us unique

NCS-FO：揭示使我们每个人独一无二的认知和神经指纹

• 批准号: 2219903
• 负责人: Tim Rogers
• 金额: $ 98.25万
• 依托单位: University of Wisconsin-Madison
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2022
• 资助国家: 美国
• 起止时间: 2022-09-01 至 2025-08-31
• 项目状态: 未结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2219903&HistoricalAwards=false
• 关键词: NCS; FO; Uncovering; cognitive; neural

Efforts to understand the brain systems that support mental abilities often assume that all people share a common cognitive and neural “template” that differs only subtly and randomly from person to person. Yet many classic studies have shown substantial variation in both mind and brain across different groups. Language relies on different brain systems for left handers compared to right handers. Students who struggle with reading show different patterns of neural connectivity than do fluent readers. Math experts see different kinds of relationships amongst even simple integers than do college undergraduates. This project will combine new computational insights, functional brain imaging, and brain stimulation to characterize variability in the mental and neural structures that support perception, language, social cognition, and high-order thought. This work has potential to impact human health and patterns of recovery from brain damage, approaches to education for typical and atypical populations, and other fields of science that can benefit from discovery of patterned variation.The scientific goals involve developing new theory and algorithms for signal discovery in behavioral and neural data, and using these methods to identify commonalities and differences across individuals in the structure of their mental representations and the corresponding neural underpinnings. For both brain and behavior, structure estimation requires a lot of data, and data generated by a single participant is typically noisy. To address this, classic approaches treat data from each participant as a sample drawn from an underlying system common to all members of a population. In effect, individual differences are treated as noise. Our approach instead assumes that individuals can differ significantly in the mental and neural structures that generate data, but that such variation is structured in ways that can aid in signal-discovery. This project will formalize these ideas in a computational approach to structure estimation, establish bounds on sample complexity for the model, and develop practical algorithms for applying the approach to behavioral and brain imaging data. In parallel with this computational stream, the investigators will collect behavioral data, functional brain images, and brain-stimulation data for several tasks tapping different aspects of cognition. These experiments, combined with the new computational methods, will experimentally determine how people differ in the structure of their mental and neural representations—potentially changing our basic understanding of the different ways brains can give rise to minds. This project is funded by the Integrated Strategies for Understanding Neural and Cognitive Systems (NCS) program, which is jointly supported by the Directorates for Computer and Information Science and Engineering (CISE), Education and Human Resources (EHR), Engineering (ENG), and Social, Behavioral, and Economic Sciences (SBE).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.

理解支持智力的大脑系统的努力通常假设所有人都有一个共同的认知和神经“模板”，这个模板只是在人与人之间有细微和随机的不同。然而，许多经典研究表明，不同的群体在思维和大脑方面都存在很大差异。与右撇子相比，左撇子的语言依赖于不同的大脑系统。与阅读流利的学生相比，阅读困难的学生表现出不同的神经连接模式。数学专家认为，即使是简单的整数之间也存在不同类型的关系，而不是大学本科生。这个项目将结合新的计算洞察力、脑功能成像和脑刺激来表征支持感知、语言、社会认知和高阶思维的精神和神经结构的可变性。这项工作有可能影响人类的健康和脑损伤的恢复模式，典型和非典型人群的教育方法，以及其他可以从模式变异的发现中受益的科学领域。科学目标包括开发新的理论和算法，用于在行为和神经数据中发现信号，并使用这些方法来识别不同个体在心理表征结构和相应的神经基础上的共性和差异。对于大脑和行为来说，结构估计都需要大量的数据，而单个参与者产生的数据通常是有噪声的。为了解决这一问题，经典方法将来自每个参与者的数据视为从人群中所有成员共同的基础系统中提取的样本。实际上，个体差异被视为噪音。相反，我们的方法假设，在产生数据的心理和神经结构上，个体可以有很大的不同，但这种差异的结构有助于信号发现。该项目将在结构估计的计算方法中将这些想法形式化，为模型建立样本复杂性的界限，并开发将该方法应用于行为和大脑成像数据的实用算法。与这一计算流并行，研究人员将收集行为数据、脑功能图像和脑刺激数据，用于几个挖掘认知不同方面的任务。这些实验与新的计算方法相结合，将在实验上确定人们在精神和神经表征结构上的不同--潜在地改变我们对大脑产生思维的不同方式的基本理解。本项目由理解神经和认知系统的综合策略(NCS)计划资助，该计划由计算机和信息科学与工程(CEISE)、教育和人力资源(EHR)、工程(ENG)和社会、行为和经济科学(SBE)等主管部门共同支持。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

项目成果

One for All: Simultaneous Metric and Preference Learning over Multiple Users

• DOI: 10.48550/arxiv.2207.03609
• 发表时间: 2022-07
• 期刊: ArXiv
• 作者: Gregory H. Canal;Blake Mason;Ramya Korlakai Vinayak;R. Nowak

Efficiently Learning Relative Similarity Embeddings with Crowdsourcing

• DOI: 10.21105/joss.04517
• 发表时间: 2023-04
• 期刊: J. Open Source Softw.
• 作者: Scott Sievert;R. Nowak;Timothy Rogers

Decoding semantic representations in mind and brain

• DOI: 10.1016/j.tics.2022.12.006
• 发表时间: 2023-01
• 期刊: Trends in Cognitive Sciences
• 影响因子: 19.9
• 作者: Saskia L. Frisby;A. Halai;Christopher R. Cox;M. L. Lambon Ralph;T. Rogers