Group housing: A robotic system to track and interact with individuals

描述（由申请人提供）：已知单独圈养动物，即使是短时间，也会对行为、学习能力和生理产生负面影响。群体圈养的啮齿动物能够通过改善行为和生理反应更好地科普压力。不幸的是，由于技术限制，许多实验，包括设计用于测量行为、昼夜节律活动和/或神经学研究的那些实验，都需要单个外壳。实验性干预的自动传递，例如刺激，食物奖励或温和的轻推以扰乱睡眠，很容易在单个动物身上完成，但在群体环境中很复杂。本提案的总体目标是（1）开发一种新型笼式围栏，其设计用于分组圈养多只身体相似的动物，并具有长期的计算机辅助视频监控和射频识别（RFID），以正确和准确地识别和跟踪四只动物一组中的个体动物， (2)以自动地向组内的一个或多个动物施加干预刺激。该系统将采用位于笼架正上方的机器人臂，当该机器人臂与视频跟踪模块和RFID读取器连接时，可以选择性地施加干预刺激（即，温柔的轻推、食物、空气喷吹）到特定的动物。可以使用经由视频跟踪、无线脑电图（EEG）测量、特定神经化学物质的无线生物传感器测量或其他无线生理测量收集的数据来对所定义的状态进行分类。该系统将是可配置的，使得干预刺激可以基于由研究者定义的生理参数被施加到个体动物。在第一阶段，Pinnacle成功设计并测试了一个原型系统，该系统在一组四只动物中识别出一只动物，并通过轻推6小时剥夺了该动物的睡眠。成功地测量了睡眠反弹，证明了该系统的可行性。第二阶段将通过测试和改进涉及多只动物的动物跟踪和行为，提高视频跟踪的准确性，并通过创建可以轻松互换的多个手臂附件来增加研究人员可用的干预刺激类型，继续这一发展。在第二阶段完成后，我们将有一个完整的，交钥匙的硬件和软件解决方案，将商业上提供给研究界。

DESCRIPTION (provided by applicant): Individually housing animals, even for a short period, is known to have a negative influence on behavior, learning ability, and physiology. Group-housed rodents are better able to cope with stress through improved behavioral and physiological responses. Unfortunately, many experiments, including those designed to measure behavior, circadian activity, and/or neurological studies, necessitate single- housing due to technological restrictions. Automated delivery of an experimental intervention, such as a stimulus, a food reward or gentle nudging to disturb sleep, is easily accomplished with single animals but is complicated in group-housed environments. The overall goal of this proposal is (1) to develop a novel cage enclosure designed to group house multiple, physically similar animals with long-term computer-aided video monitoring and radio frequency identification (RFID) to correctly and accurately identify and track individual animals within a group of four and (2) to automatically apply an intervention stimulus to one or more animals within the group. This system will employ a robotic arm situated directly above the cage, which, when tied to the video tracking module and RFID reader, can selectively apply an intervention stimulus (i.e., gentle nudging, food, air puff) to a specific animal based on state- specific feedback. The defined state may be classified using data collected via video tracking, wireless electroencephalograph (EEG) measurement, wireless biosensor measurement of specific neurochemicals, or other wireless physiological measurements. The system will be configurable so that the intervention stimulus may be applied to an individual animal based on physiological parameters defined by the researcher. In Phase I, Pinnacle successfully designed and tested a prototype system that identified one animal in a cohort of four and sleep deprived that animal by nudging for 6 hours. Sleep rebound was successfully measured to prove feasibility of the system. Phase II will continue this development by testing and refining animal tracking and behavior involving multiple animals, increasing the accuracy of video tracking, and increasing the types of intervention stimuli available to the researcher by creation of multiple arm attachments that can be easily interchanged. At the completion of Phase II, we will have a complete, turn-key hardware and software solution that will be commercially available to the research community.