Tools for assessing shared sensory experience in interacting mice.

目的：从长远来看，我们围绕大脑在特定回路中的活动与动物行为相联系的机制开发了新的方法和知识。在接下来的5年里，我们扩展了我们的新范式，在量化它们的行为特征的同时，同时评估两只小鼠的大脑活动。我们的数据使我们将重点放在老鼠胡须系统作为行为交互的关键感觉通道上。我们认为，小鼠之间发生的相互胡须接触可能为受试者之间的大脑活动提供联系的信号。我们的目标是了解相互胡须感觉体验和受试者间大脑活动夹带之间的联系。为了同时实现这一目标，我们开发了新的行为分析工具，这些工具是由合成老鼠模型实现的。我们的方法使用自动视频分析工具报告鼠标行为的各个方面，这些工具是通过合成数据增强策略进行训练的。这项发现号5年研究计划将使用合成数据增强和新的脑成像工具来利用关于小鼠行为交互过程中共享的感觉胡须体验的信息。然而，还没有研究追踪两只相互作用的小鼠的所有胡须。我们将使用一只合成的动画老鼠来训练下一代算法，以同时跟踪所有胡须；该模型还将评估诸如瞳孔直径和运动等面部特征。这种综合方法为模型训练提供了更大的灵活性，因为胡须的身份是从地面真实的模型依赖数据中得知的，并且可以模拟更精确的瞳孔动态，从而更好地与正在进行的大脑活动相关。重要性：研究对象之间共享经验的神经基础是一个重要的悬而未决的问题。人类受试者的大脑间同步性已经被研究过，但很少有动物模型存在。我们的方法是使用双重大脑成像和小鼠之间受限的行为交互来识别独特的大脑活动特征。对相互胡须运动的检查，即小鼠通过交替的扭动相互探测，是与受试者之间高度的大脑同步性相关的行为特征。在更大的范围内，我们测试合成动物提供训练数据以解决相关生物学问题的能力。这些研究支持了我们的长期目标，即探索动物行为特定特征背后的大脑回路。影响：我们为神经科学带来新的计算工具的工作在领域内和加拿大都产生了影响。我们提供了关于潜在行为互动的大脑电路的新知识，以及与高比例正在进行的大脑活动密切相关的面部特征(胡须和瞳孔)的详细合成老鼠模型。此外，我们还将培训下一代科学家，在丰富和互动的环境中参与所有级别的HQP，结合物理科学方法和与基本动物行为的联系。

Objectives: Over the long term we develop new approaches and knowledge around mechanisms by which brain activity in specific circuits is linked to animal behaviors. Over the next 5 years we extend our novel paradigms which assess the brain activity of two mice simultaneously while quantifying features of their behavior. Our data have led us to focus on the mouse whisker system as a key sensory modality for behavioral interaction. We propose that mutual whisker contact that occurs between mice may provide a signal for linking brain activity between subjects. We aim to understand the link between mutual whisker sensory experience and entrainment of inter-subject brain activity. To accomplish this goal in parallel we develop novel behavioral analysis tools that are enabled by a synthetic mouse model. Our methods report aspects of mouse behavior using automated video analysis tools that are trained using a synthetic data augmentation strategy. This Discovery 5 year research program will use synthetic data augmentation and new brain imaging tools to leverage information about shared sensory whisker experience during mouse behavioral interaction. However, no study has tracked all whiskers in 2 interacting mice. We will use a synthetic animated mouse to train next generation algorithms to track all whiskers simultaneously; the model will also evaluate facial features such as pupil diameter and movement. This synthetic approach provides greater flexibility for model training since whisker identity is known from ground truth model-dependent data and more precise pupil dynamics can be simulated and thus better related to on-going brain activity. Importance: The neural basis of shared experiences between subjects represents an important open question. Inter-brain synchrony has been studied in human subjects, but few animal models exist.  Our approach is to use dual brain imaging and a constrained behavioral interaction between mice to identify features of brain activity that are unique. Examination of mutual whisker movements where mice probe each other with alternating bouts of whisking is a behavioral signature associated with high inter-subject brain synchrony. On a larger scale, we test the ability of synthetic animals to provide training data to address relevant biological questions. These studies support our long term goal of probing the brain circuits underlying specific features of animal behaviour. Impact: Both within-field and Canadian impact of our work that brings novel computing tools to neuroscience. We provide new knowledge on brain circuitry underlying behavioral interaction as well as detailed synthetic mouse models for facial features (whiskers and the pupil) that are tightly linked to a high-proportion of on-going brain activity. As well, we will train the next generation of scientists engaging HQPs on all levels within a rich and interactive setting that incorporates physical science approaches and links to basic animal behavior.