ASM: Multi-Sensory Control of Tracking Behavior in Weakly Electric Fish

项目摘要asm：弱电鱼跟踪行为的多感官控制，pi: Noah J. Cowan，机械工程系，约翰霍普金斯大学共同pi: Eric S. Fortune，心理与脑科学系，约翰霍普金斯大学智力价值-一个紧密集成的多学科方法，使用一个独特适合的模型系统，将有助于回答感觉运动整合的一个基本问题：神经系统是如何处理信息来控制运动的？在模型系统中，弱电鱼稳健而自然地来回游动以稳定视觉和/或电感觉图像，就像人类用眼睛平滑地跟踪运动物体以稳定视觉图像一样。这项合作工作建立在PI（感觉运动系统建模师）和Co-PI（有机体感觉神经生物学家）的优势之上。该方法结合了数学建模、行为实验和神经生理学分析。跟踪行为的数学模型可以对行为和神经处理进行具体的、可测试的预测。该模型的行为预测将通过系统地改变视觉和电感觉信息来测试，这些信息可用于跟踪移动的感觉图像。该模型的预测也将通过中枢神经系统记录在清醒、行为正常的动物身上进行测试。刺激将包括与用于行为实验的信号相同的信号，其输入-输出关系可以从模型中预测。本科生和研究生将接受神经科学、实验设计、数据收集和分析以及计算建模等多学科培训。此外，我们将共同教授一门新的基于传感器的动物运动的本科课程，以及一个配套的研究生研讨会。最后，对感觉运动动物行为的研究具有启发人工系统自主控制新策略的巨大潜力。

Project AbstractASM: Multi-Sensory Control of Tracking Behavior in Weakly Electric FishPI: Noah J. Cowan, Dept. of Mechanical Engineering, Johns Hopkins UniversityCo-PI: Eric S. Fortune, Dept. of Psychological and Brain Sciences, Johns Hopkins UniversityIntellectual Merit- A tightly integrated and multidisciplinary approach using a uniquely suited model system will help answer a fundamental question in sensorimotor integration: how is information processed by the nervous system to control locomotion?In the model system, weakly electric fish robustly and naturally swim back and forth to stabilize visual and/or electrosensory images, just as humans smoothly track moving objects with their eyes to stabilize visual images. This collaborative work builds on the strengths of the PI, a modeler of sensorimotor locomotion systems and the Co-PI, an organismal sensory neurobiologist.The approach incorporates mathematical modeling, behavioral experiments, and neurophysiological analyses. A mathematical model of the tracking behavior makes specific, testable predictions of both behavior and neural processing. The model's predictions of behavior will be tested by systematically varying visual and electrosensory information available to the fish for tracking a moving sensory image. The model's predictions will also be tested using central nervous system recordings in awake, behaving animals. The stimuli will include signals identical to those used for behavioral experiments, whose input-output relations are predictable from the model. Broader Impacts Undergraduate and graduate trainees will receive multidisciplinary training in neuroscience, experimental design, data collection and analysis, and computational modeling. Further, we will co-teach a new undergraduate course in sensor-based animal locomotion, and a companion graduate seminar. Finally, the study of sensorimotor animal behaviors has greatpotential to inspire novel strategies for autonomous control in artificial systems.