Integrating Human Wearers' Perception and Cognition into Prosthesis Control Policy

这项拨款将支持研究，将贡献新的知识与人-假体相互作用和机器人假体控制的个性化。目前的动力下肢机器人假肢利用关节运动和力的内在反馈来调节关节阻抗，作为当前步态阶段的函数。控制参数的个性化（控制器调整）通常由临床医生手动和启发式地执行，每次修改一个参数，直到截肢者的步态“看起来不错”，并且截肢者自我报告对控制的满意度。该项目的研究目标是为LL机器人假体开发一种新型的“佩戴者主导”自动调整程序，该程序考虑并增强步态过程中人机交互的认知方面，如用户的目标、所需的注意力、认知工作量、信任和舒适度。本项目将通过开发智能假肢控制器，根据用户的个人偏好和感知进行调整，从而提高下肢截肢患者的身体表现，最大限度地接受机器肢体作为身体的功能部分，从而促进科学进步，促进国民健康。该项目的更广泛影响包括在本科和研究生阶段的K-12推广和教育工作，旨在吸引和留住女性和代表性不足的少数民族进入STEM领域。该项目包括三组研究活动，旨在为下肢机器人假体开发一种用户自适应自动调谐程序。第一项研究开发了一种新颖的“穿戴式”调节程序，其中身体健全的受试者和单侧经股截肢的个体将与机器人假体一起行走，并使用“出声思考”过程调节其控制参数。研究小组将进行人体实验，研究佩戴者在使用机器人肢体行走时的思维和决策过程，并调整其控制参数以优化其性能和舒适度。第二组比较了假肢行走时的身体步态表现和认知功能的选定方面，这些假肢由现有的机器学习算法、训练有素的假肢师或使用新型佩戴者主导的调整程序进行调整。生物力学性能和知觉测量、认知负荷、注意力分配和其他认知因素将在三种调谐方法中进行比较。研究结果将在认知和物理层面填补穿戴者与机器人交互的知识空白。第三组活动将把新的由佩戴者主导的调节过程整合到一个新的、智能的、自动化的下肢假肢调节系统中，该系统直接将佩戴者的偏好、感知和认知融入到假肢个性化的过程中，从而促进佩戴者与假肢的无缝集成和实施。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

This grant will support research that will contribute new knowledge related to human-prosthesis interactions and the personalization of robotic prosthesis control. Current powered lower limb (LL) robotic prostheses use intrinsic feedback of joint motion and forces to adjust joint impedance as a function of the current phase of gait. Personalization of the control parameters (controller tuning) is typically performed manually and heuristically by a clinician who modifies one parameter at a time until the amputee's gait "looks good" and the amputee self-reports satisfaction with the control. The research objective of this project is to develop a novel "wearer-led" auto-tuning procedure for LL robotic prostheses that considers and enhances cognitive aspects of the human-machine interaction during gait, such as the user's goals, required attention, cognitive workload, trust and comfort. This project will promote the progress of science and advance the national health by developing intelligent prosthesis controllers that can tune themselves to the personal preferences and perceptions of their users, thereby augmenting the physical performance of individuals with lower limb amputations and maximize their acceptance of the robotic limbs as a functional part of the body. Broader impacts of the project include K-12 outreach and educational efforts at the undergraduate and graduate levels intended to attract and retain women and underrepresented minorities into STEM fields.The project involves three sets of research activities aimed at developing a user-adaptive auto-tuning procedure for lower limb robotic prostheses. The first develops a novel "wearer-led" tuning procedure wherein able-bodied subjects and individuals with unilateral trans-femoral amputation will walk with the robotic prosthesis and tune its control parameters using a "think aloud" process. The research team will perform human subject experiments to study the wearers' thought and decision processes as they walk with a robotic limb and adjust its control parameters to optimize its performance and comfort. The second set compares physical gait performance and selected aspects of cognitive functioning while walking with a prosthesis that is tuned either by an existing machine learning algorithm, by a trained prosthetist, or using the novel wearer-led tuning procedure. Biomechanical performance and measures of perception, cognitive workload, attentional allocation, and other cognitive factors will be compared across the three tuning approaches. The results will fill the knowledge gap in wearer-robot interactions at the cognitive and physical levels. The third set of activities will incorporate the novel wearer-led tuning procedure into a new, intelligent, automated, lower limb prosthesis tuning system that directly incorporates the wearer's preference, perception, and cognition into the procedure for prostheses personalization, thereby promoting seamless wearer-prosthesis integration and embodiment.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.