Predictability in complex object control

复杂对象控制的可预测性

• 批准号: 10365518
• 负责人: Dagmar Sternad
• 金额: $ 68.53万
• 依托单位: NORTHEASTERN UNIVERSITY
• 依托单位国家: 美国
• 财政年份: 2015
• 资助国家: 美国
• 起止时间: 2015-09-24 至 2027-01-31
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10365518
• 关键词: Address; Attenuated; Basic Science; Behavior; Clinic; Clinical; Coffee; Complex; Custom; Device Designs; Devices; Dimensions; Ensure; Environment; Failure; Feasibility Studies; Feedback; Female; Funding; General Hospitals; Goals; Hand; Human; Impairment; Individual; Information Theory; Lead; Learning; Life; Manuals; Massachusetts; Measurement; Methodology; Modeling; Modification; Motor; Movement; Neurologic; Neurosciences; Noise; Nonlinear Dynamics; Patients; Performance; Periodicity; Process; Property; Recovery; Research; Research Personnel; Risk; Robotics; Severities; Stroke; System; Testing; Translating; Upper Extremity; base; clinical application; cohort; computational neuroscience; dexterity; dynamic system; experience; experimental study; functional disability; haptics; high dimensionality; innovation; insight; internal control; medical schools; motor control; motor function recovery; motor impairment; motor recovery; multidisciplinary; nervous system disorder; novel; novel strategies; object motion; portability; post stroke; preempt; prevent; relating to nervous system; scale up; sensorimotor system; simulation; stroke patient; tool; virtual; virtual environment; virtual human

PROJECT SUMMARY Manipulation of complex objects or tool use is a hallmark of daily living, and loss of manual dexterity due to motor impairments lead to loss of independence. Manipulating objects is particularly challenging when the object has internal dynamics that is not directly controlled. Even the seemingly simple task of transporting a cup of coffee has intrinsic dynamics that humans need to predict, preempt, and compensate for to avoid spilling. Control of such complex nonlinear systems with online error corrections based on precise internal models appears daunting, given the slow neural processes and the ubiquitous noise in the sensorimotor system. Hence, this research tests the hypothesis that humans learn to simplify the object interactions, i.e., make the interactions predictable. The task of carrying a cup of coffee is modeled with a cart-and-pendulum system that is rendered in a virtual environment and subjects interact with the virtual cup via a robotic manipulandum. To gain insight into human control strategies, this proposal develops a task-dynamic approach that affords principled hypothesis-testing by parsing the complex dynamics into execution and result variables, with minimal assumptions about the human controller. Eight experiments test the overall hypothesis that humans seek solutions that are predictable, by correlating hand-object motions, and making the behavior stable and tolerant to error and risk to obviate error corrections and prevent failure. Aim-1 tests control of internal dynamics in linear movements and examines how humans choose initial conditions to mitigate perturbations, how they preempt undesired ball oscillations, how they exploit intermittent contact to develop a stable rhythm, and how they modify the object properties to facilitate stable contact behavior. To examine learning, Aim-2 scales up the dimensionality of the task by introducing more real-life planar cup movements, which creates an exponential increase in complexity. Four experiments test task goals that introduce new dynamic challenges, such as combination of rhythmic and discrete movements, complex ball dynamics when changing movement directions, adaptation and modification of object properties, all to show how humans either exploit or override internal dynamics to achieve predictability. Aim-3 introduces a real version of the task with a custom-designed device, the MAGIC Table. Following a comparison of the real and virtual set-ups, the MAGIC Table is used to leverage the theoretical framework to create novel sensitive metrics to quantify motor function for clinical applications. Specifically, we assess severity and recovery of motor impairment in a cohort of patients after stroke. As manual dexterity is compromised in many individuals with neurological disorders, the experimental paradigm and its quantitative analyses promise to become a useful platform to gain insights into neurological diseases.

项目摘要 复杂物体的操作或工具的使用是日常生活的标志，并且由于运动而失去手动灵活性。 残疾导致丧失独立性。当对象具有以下特性时，操纵对象尤其具有挑战性： 不直接控制的内部动态。即使是看似简单的运送一杯咖啡的任务 人类需要预测、抢占和补偿以避免溢出的内在动力。控制 基于精确内模的在线误差修正的复杂非线性系统 考虑到缓慢的神经过程和感觉运动系统中无处不在的噪音，这是令人生畏的。所以这 研究检验了这样的假设：人类学会了简化对象交互，即，使互动 可预测的。携带一杯咖啡的任务是用一个手推车和钟摆系统建模的， 在虚拟环境中，受试者通过机器人操纵器与虚拟杯子交互。获得洞察力 人类控制策略，这项建议开发了一个任务动态的方法，提供原则性的 通过将复杂的动态解析为执行和结果变量来进行假设测试， 关于人类控制者的假设。八个实验验证了人类寻求 解决方案是可预测的，通过关联手-物体运动，并使行为稳定和宽容 错误和风险，以纠正错误并防止失败。Aim-1测试线性系统中内部动态的控制 运动和研究人类如何选择初始条件，以减轻扰动，他们如何抢占 不受欢迎的球振荡，他们如何利用间歇性接触来发展稳定的节奏，以及他们如何修改 对象属性以促进稳定的接触行为。为了检查学习，Aim-2将 通过引入更多现实生活中的平面杯运动，这创造了一个指数， 增加复杂性。四个实验测试引入新的动态挑战的任务目标，例如 节奏和离散运动的组合，改变运动方向时的复杂球动力学， 适应和修改对象属性，所有这些都表明人类如何利用或重写内部属性， 动态实现可预测性。Aim-3通过定制设计的设备引入了该任务的真实的版本， 魔法桌子在比较了真实的和虚拟的设置之后， 理论框架，以创建新的敏感指标，以量化运动功能的临床应用。 具体来说，我们评估了一组中风患者运动功能障碍的严重程度和恢复情况。人工 灵活性在许多患有神经系统疾病的个体中受到损害，实验范例及其 定量分析有望成为深入了解神经系统疾病的有用平台。