Adaptive Smart Controller For Brain-Prosthetic Hand Interface

用于脑假手接口的自适应智能控制器

• 批准号: 7671272
• 负责人: Samuel Thomas Clanton
• 金额: $ 4.62万
• 依托单位: CARNEGIE-MELLON UNIVERSITY
• 依托单位国家: 美国
• 财政年份: 2007
• 资助国家: 美国
• 起止时间: 2007-09-01 至 2012-08-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/7671272
• 关键词: Algorithms; Behavior; Brain; Computer software; Data; Development; Devices; Elements; Ensure; Environment; Fingers; Freedom; Goals; Hand; Human; Injury; Lead; Life; Limb Prosthesis; Measures; Motor Cortex; Neurodegenerative Disorders; Physiological; Population; Positioning Attribute; Primary Lateral Sclerosis; Primates; Prosthesis; Psychophysiology; Public Health; Research; Robot; Robotics; Shapes; Signal Transduction; Spinal cord injury; System; Testing; Therapeutic; Time; United States; Upper Extremity; Upper arm; Work; clinical application; computerized data processing; flexibility; functional restoration; grasp; improved; kinematics; loss of function; nervous system disorder; neural prosthesis; neuroregulation; patient population; relating to nervous system; research study; vector

DESCRIPTION (provided by applicant): Over 130,000 people in the United States live with partial or complete loss of function of the arm and hand due to injury or neurodegenerative disease. Neural prosthetic approaches could lead to therapeutic devices that have the potential to restore function and independence to this patient population. Our work is in neural prosthetic interfaces that process signals from the motor cortex to control robotic arms and hands. While much progress has been made in the neural control of a prosthetic arm, the complexities inherent in the control of the hand make the development of a neural prosthetic hand interface difficult. Current work is in the development of a neural prosthetic robot hand controller. Our aim is that it will have the flexibility to adapt to variable neural control signals while employing an automated grasp planning system to ensure stable grasping. This system will act as a bridge to complete neural control of the hand while having immediate clinical application in improving current upper limb prostheses. Specific Aim 1: To develop a robotic neural prosthetic hand control system that dynamically integrates elements of a sophisticated automated grasp planning algorithm with grasp-related information derived from neural firing in the motor cortex. Data from past human studies and current primate psychophysical experiments will be used to characterize controlled features of grasping, such as kinematic coordination of the fingers and measures of manipulability of objects with achieved grasps. Recordings from the motor cortex during grasping will be used to characterize correlation of neural firing with hand position, orientation, and shaping using population vectors, an approach used currently for the neural control of a robot arm. This data will be integrated into an existing grasp planning system to focus the search space for planning successful grasping in a way that is compatible with observed physiologic behavior and neural signal-derived intent. Specific Aim 2: To test the hypothesis that a smart hand control system can allow a primate to perform cortically controlled complete reaching and grasping tasks. The grasp planning system will be integrated into an existing setup for performing on-line neural control of an upper limb prosthetic, in which primates will perform neurally controlled robotic grasping tasks. Relation to Public Health: Neural prosthetics have the potential to help people with spinal cord injury and other neurologic disorders achieve a greater degree of independence and control of their lives. Our goal is to create a control system for a neural prosthetic hand that allows natural and effective grasping.

描述(由申请人提供)： 在美国，超过13万人因受伤或神经退行性疾病而部分或完全丧失手臂和手的功能。神经假体方法可能导致治疗装置，有可能恢复这些患者的功能和独立性。我们的工作是神经假体接口，它处理来自运动皮质的信号，以控制机械臂和手。虽然在假臂的神经控制方面已经取得了很大的进展，但手的控制固有的复杂性使得神经假手接口的开发变得困难。目前的工作是开发一种神经假肢机器人手控制器。我们的目标是，它将具有灵活性，以适应不同的神经控制信号，同时采用自动抓取规划系统，以确保稳定的抓取。该系统将作为完成手的神经控制的桥梁，同时在改进现有的上肢假体方面立即应用于临床。 具体目标1：开发一种机器人神经假手控制系统，该系统动态地将复杂的自动抓取规划算法的元素与来自运动皮质神经放电的抓取相关信息集成在一起。 来自过去的人类研究和目前的灵长类心理物理实验的数据将被用来描述抓取的受控特征，例如手指的运动学协调以及具有所获得的抓取的物体的可操作性的测量。抓取过程中运动皮质的记录将被用来描述神经放电与手的位置、方向和形状的相关性，使用种群向量，这是目前用于机械臂神经控制的一种方法。这些数据将被集成到现有的抓取规划系统中，以与观察到的生理行为和神经信号派生的意图兼容的方式集中搜索空间，以规划成功的抓取。 具体目标2：验证智能手控系统可以让灵长类动物完成大脑控制的完成伸手和抓取任务的假设。 抓取规划系统将被集成到现有的设置中，用于执行上肢假肢的在线神经控制，其中灵长类动物将执行神经控制的机器人抓取任务。与公共卫生的关系：神经假体有可能帮助脊髓损伤和其他神经疾病患者实现更大程度的独立性和对他们生活的控制。我们的目标是创建一种神经假手的控制系统，使其能够自然有效地抓取。