Low-cost Tactile Sensing and Intelligent Reflexes for Prosthetic Hands

描述（由申请人提供）：肌电假手使用者在抓握和操纵易碎物体的能力方面受到严重阻碍。在这些系统中，从截肢者的残余肌肉记录的肌电图（EMG）信号被处理以产生与肌肉激活成比例的电压，从而驱动假手的DC马达。通过训练，用户可以自动打开，关闭和停止手与指尖的速度成比例控制。如果用户施加连续的EMG信号来抓握物体，则驱动指尖的电机将在物体上失速，根据EMG幅度产生30- 100 N的高抓握力。如此高的力使得在不损坏易碎物体的情况下抓住它们非常具有挑战性。对肌电信号的仔细控制和定时可以用来以非常慢的速度移动手指，在易碎的物体被压碎之前停止它们;然而，这种方法需要直视，高度集中，并且不一致。因此，大多数假肢使用者避免处理易碎的物体，限制了他们的假肢的实用性，特别是在双手任务中。在迄今为止的研究中，我们已经开发出生物启发的策略来解决这个问题。每个指尖上的触觉传感器检测何时进行接触，并调节对电机的控制信号，类似于抑制性生物反射。对于一系列低到中等振幅的EMG信号，这种反射将导致抓握闭合在低力下停止，防止易碎物体被压碎。这种方法不限制受试者产生具有较大EMG信号的较高力以抓握较重物体的能力，或者如果这是期望的结果则甚至压碎物体的能力。这种方法使用户在更大范围的抓握力上进行比例控制，几乎没有认知负担。初步研究表明，这一功能大大改善了涉及易碎或可变形物体的日常生活活动。在这项研究中，我们将开发一种低成本和强大的触觉传感器，满足商业合作伙伴提出的设计要求，并将这些传感器与商业上可用的假肢手。将与第三方临床研究合作伙伴一起开发和评估用于评估该技术和替代技术的结果指标。这项研究和开发将侧重于证明这项技术的可行性，并在商业化之前减少或消除风险。第一阶段的终点将导致一个低成本的传感器准备更大规模的集成和临床试验，以证明上层的好处 截肢者、临床医生和第三方付款人。

DESCRIPTION (provided by applicant): Myoelectric prosthetic hand users are severely hampered in their ability to grasp and manipulate fragile objects. In these systems, electromyography (EMG) signals recorded from an amputee's residual muscles are processed to produce an electrical voltage proportional to muscle activation to drive the DC motors of a prosthetic hand. With training, users can voluntarily open, close and stop the hand with proportional control over the speed of the fingertips. If the user applies a continuous EMG signal to grasp an object, the motors driving the fingertips will stall on the object, producing high grasping forces of 30-100N depending on the EMG amplitude. Such high forces make it very challenging to grasp fragile objects without damaging them. Careful control and timing of EMG signals can be used to move the fingers at very slow speeds, stopping them before a fragile object is crushed; however, this approach requires direct vision, intense concentration, and has inconsistent success. As a result, most prosthesis users avoid handling fragile objects, restricting the utility of their prosthesis, particularly in bimanual tasks. In research to date, w have developed biologically inspired strategies to solve this problem. A tactile sensor on each fingertip detects when contact is made and modulates the control signals to the motors, similar to an inhibitory biological reflex. For a range of low to medium amplitude EMG signals, this reflex will cause the grasp closure to stop at low forces, preventing fragile objects from being crushed. This approach does not restrict a subject's ability to produce higher forces with larger EMG signals for grasping heavier objects, or even to crush objects if this is the desired outcome. This approach gives the user proportional control over a wider range of grasping forces with virtually no cognitive burden. Preliminary research has demonstrated that this feature offers substantial improvement in activities of daily living that involve fragile or deformable objects. In this research we will develop a low-cost and robust tactile sensor that meets design requirements proposed by a commercial partner and integrate these sensors with a commercially available prosthetic hand. Outcome measures to evaluate this and alternative technologies will be developed and evaluated with third-party clinical research partners. This research and development will focus on demonstrating feasibility of this technology and reducing or eliminating risks prior to commercialization. Phase I endpoints will result in a low-cost sensor ready for larger scale integration and clinical trials to demonstrate benefits to upper extremity amputees, clinicians and third-party payers.