NRI: Active Tendon-Driven Orthosis for Prehensile Manipulation After Stroke

该项目为中风幸存者开发可穿戴和功能齐全的手矫形器，作为辅助和训练设备。对于日常生活，这样的设备可以帮助完成操作任务，增加控制感和独立生活的能力。对于康复，可穿戴矫形器可以实现现实世界操作任务的功能培训，将培训扩展到在临床环境中进行的少量会话，并在日常活动过程中提供分布式培训，而不是在指定的治疗会话中进行闭塞培训。从医疗保健的角度来看，这项工作将使在医疗监督下的诊所使用的设备向日常生活中使用的非侵入性功能矫形器转变。从工程角度来看，它将迈向一种新的主动可穿戴系统，可以通过自然运动或类似的直观界面进行控制。需要新的方法来建造一种同时灵活(对于复杂的操作)、非侵入性(用于日常生活)和直观控制的设备。为了实现这一点，我们研究了使用在手和手臂表面布线的人造肌腱网络来帮助关节移动的方法，而不增加大型结构组件，如外骨骼。我们方法的一个关键原则是，手矫形器可以在日常操作任务中提供有意义的帮助，即使使用的致动器数量远远少于手中关节的数量。我们还将研究多种控制方法，包括单手手术(检测并增强瘫痪肢体的内源性运动)和双侧手术(利用受影响较小的肢体的运动为瘫痪肢体产生控制信号)。

This project develops wearable and functional hand orthoses for stroke survivors, to act as both assistive and training devices. For daily living, such a device could provide assistance with manipulation tasks, increasing the sense of control and the ability to live independently. For rehabilitation, a wearable orthosis could enable functional training on real-world manipulation tasks, extend training beyond the small number of sessions performed in a clinical setting, and provide distributed training during the course of daily activities rather than block training in designated therapeutic sessions. From a healthcare perspective, this work will enable a shift from devices used in a clinic under medical supervision towards unobtrusive functional orthoses used during daily life. From an engineering perspective, it will advance towards a new class of active wearable systems, controllable by natural movement or similarly intuitive interfaces. New approaches will be required to build a device that is simultaneously dexterous (for complex manipulation), non-intrusive (for daily living) and intuitive to control. To achieve this, we investigate ways to assist joint movement using a network of artificial tendons routed on the surface of the hand and arm, without adding large structural components such as an exoskeleton. A key tenet of our approach is that a hand orthosis can provide meaningful assistance with daily manipulation tasks even when using a number of actuators far smaller than the number of joints in the hand. We will also investigate multiple control methods, including unimanual operation (where endogenous movement of the paretic limb is detected and enhanced) and bilateral operation (where movement of the less affected limb is used to generate control signals for the paretic one).