Davoodi Rahman, Urata Chet, Hauschild Markus, Khachani Mehdi, Loeb Gerald E
Alfred Mann Institute for Biomedical Engineering, Department of Biomedical Engineering, University of Southern California, 1042 Downey Way, Los Angeles, CA 90089, USA.
IEEE Trans Biomed Eng. 2007 Nov;54(11):1909-18. doi: 10.1109/TBME.2007.902252.
Neural prostheses for restoration of limb movement in paralyzed and amputee patients tend to be complex systems. Subjective intuition and trial-and-error approaches have been applied to the design and clinical fitting of simple systems with limited functionality. These approaches are time consuming, difficult to apply in larger scale, and not applicable to limbs under development with more anthropomorphic motion and actuation. The field of neural prosthetics is in need of more systematic methods, including tools that will allow users to develop accurate models of neural prostheses and simulate their behavior under various conditions before actual manufacturing or clinical application. Such virtual prototyping would provide an efficient and safe test-bed for narrowing the design choices and tuning the control parameters before actual clinical application. We describe a software environment that we have developed to facilitate the construction and modification of accurate mathematical models of paralyzed and prosthetic limbs and simulate their movement under various neural control strategies. These simulations can be run in real time with a stereoscopic display to enable design engineers and prospective users to evaluate a candidate neural prosthetic system and learn to operate it before actually receiving it.
用于恢复瘫痪和截肢患者肢体运动的神经假体往往是复杂的系统。主观直觉和试错方法已应用于功能有限的简单系统的设计和临床适配。这些方法耗时、难以大规模应用，且不适用于具有更多拟人化运动和驱动的正在研发的肢体。神经假肢领域需要更系统的方法，包括能让用户在实际制造或临床应用之前开发神经假肢的精确模型并模拟其在各种条件下行为的工具。这种虚拟原型将提供一个高效且安全的试验台，以便在实际临床应用之前缩小设计选择范围并调整控制参数。我们描述了一个我们开发的软件环境，以促进瘫痪和假肢肢体精确数学模型的构建和修改，并模拟它们在各种神经控制策略下的运动。这些模拟可以通过立体显示实时运行，使设计工程师和潜在用户能够在实际接收之前评估候选神经假体系统并学习如何操作它。
