CAREER: Hierarchical Representations for Visual Categorization and Decision Making

职业：视觉分类和决策的层次表示

• 批准号: 0955640
• 负责人: David Freedman
• 金额: $ 96.33万
• 依托单位: University of Chicago
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2010
• 资助国家: 美国
• 起止时间: 2010-07-01 至 2015-06-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0955640&HistoricalAwards=false
• 关键词: CAREER; Hierarchical; Representations; Visual; Categorization

Humans and other advanced animals have an impressive capacity to recognize the behavioral significance, or category membership, of a wide range of sensory stimuli. This ability is critical, because it allows us to respond appropriately to the continuous stream of stimuli and events that we encounter in our interactions with the environment. Of course, we are not born with a built-in library of meaningful categories, such as 'tables' and 'chairs,' that we are pre-programmed to recognize. Instead, we learn to recognize the meaning of such stimuli through experience. With National Science Foundation Funding, Dr. David J. Freedman is carrying out studies whose goal is to understand how visual-feature encoding in early visual processing areas is transformed into more meaningful representations at more advanced neuronal processing stages in the brain. The goals of the proposed studies are to compare neuronal representations of visual-motion processing stages across a network of interconnected brain areas in and around the parietal lobe during visual motion categorization tasks. Specifically, one series of experiments compares neuronal responses in two distinct interconnected regions of parietal cortex, the lateral and medial interparietal areas, which are known to be more involved in visual and somatosensory or motor processing, respectively. Activity in these two areas is examined during a categorization task that requires motor decisions to be executed in response to visual stimuli, allowing the relative roles of the two areas in the decision making process to be determined. A second series of experiments is comparing cortical activity in the lateral intraparietal and prefrontal cortices during a novel visual categorization task in which subjects learn multiple independent category rules and apply those rules flexibly and dynamically to incoming visual stimuli. This study gives critical insights into the contributions of frontal and parietal cortex to flexible rule-based categorization. Together, these studies can yield important insights into how learning influences the encoding of visual information and into the roles of interconnected networks of parietal and frontal cortices in visual recognition and decision making.While much is known about how the brain processes simple sensory features (such as color, orientation, and direction of motion), less is known about how the brain learns and represents the meanings or category of stimuli. A greater understanding of visual learning and categorization is critical for addressing a number of brain diseases and conditions (e.g., stroke, Alzheimer's disease, attention deficit disorder, and schizophrenia) that leave patients impaired in everyday tasks that require visual learning, recognition, and/or evaluating and responding appropriately to sensory information. Dr. Freedman's research is helping to guide the next generation of treatments for these brain-based diseases and disorders by helping to develop a detailed basic understanding of the brain mechanisms that underlie learning, memory and recognition. These studies also have relevance for understanding and addressing learning disabilities, such as attention deficit disorder and dyslexia, which affect a substantial fraction of school age children and young adults. A more detailed understanding of the basic brain mechanisms underlying learning, memory and attention will likely give important insights into the causes and potential treatments for disorders involving these cognitive faculties.

人类和其他高级动物具有令人印象深刻的能力，能够识别各种感觉刺激的行为意义或类别成员关系。这种能力是至关重要的，因为它使我们能够对我们在与环境互动中遇到的源源不断的刺激和事件做出适当的反应。当然，我们与生俱来就没有一个内置的有意义的类别库，比如我们预先编程识别的“桌子”和“椅子”。取而代之的是，我们学会通过体验来认识这些刺激的意义。在国家科学基金会的资助下，大卫·J·弗里德曼博士正在进行研究，其目标是了解早期视觉处理区域的视觉特征编码是如何在大脑中更高级的神经元处理阶段转化为更有意义的表征的。这项研究的目的是比较视觉运动分类任务中顶叶及其周围相互连接的脑区网络中视觉运动加工阶段的神经元表征。具体地说，一系列实验比较了顶叶皮质两个不同的相互关联区域--外侧和内侧顶叶间区--的神经元反应，这两个区域分别被认为更多地参与视觉、躯体感觉或运动处理。这两个区域的活动在一项分类任务中被检查，该任务要求对视觉刺激执行运动决策，从而确定这两个区域在决策过程中的相对角色。第二个系列实验是比较在一项新的视觉分类任务中，被试学习多个独立的分类规则，并灵活地、动态地将这些规则应用于传入的视觉刺激的情况下，大脑外侧、顶内和前额叶皮质的活动。这项研究对额叶和顶叶皮质对灵活的基于规则的分类的贡献给予了关键的见解。总而言之，这些研究可以对学习如何影响视觉信息的编码以及顶叶和额叶皮质相互连接的网络在视觉识别和决策中的作用产生重要的见解。虽然人们对大脑如何处理简单的感觉特征(如颜色、方向和运动方向)知道得很多，但对大脑如何学习和代表刺激的含义或类别知之甚少。更好地了解视觉学习和分类对于解决许多大脑疾病和状况(例如，中风、阿尔茨海默病、注意力缺陷障碍和精神分裂症)至关重要，这些疾病和条件会使患者在需要视觉学习、识别和/或适当评估和响应感官信息的日常任务中受损。弗里德曼博士的研究帮助人们对构成学习、记忆和认知的大脑机制有了详细的基本了解，从而有助于指导下一代针对这些脑部疾病和障碍的治疗。这些研究还与了解和解决学习障碍有关，如注意力缺陷障碍和阅读障碍，这些障碍影响到相当一部分学龄儿童和年轻人。更详细地了解学习、记忆和注意力背后的基本大脑机制，可能会为涉及这些认知能力的疾病的原因和潜在治疗方法提供重要的见解。