Adaptive Prosthetic Limbs Enhanced by Machine Learning and Soft Robotics

**“人体会改变形状，但假肢插座不会。”**与早晨相比，我们的身体在晚上大约矮1.5厘米。我们的脚和手在晚上更大。我们大多数人没有注意到这些变化，因为我们穿着灵活的衣服。但是，如果你失去了一条腿，你必须戴上一个固定在你身体上的假肢，这个假肢有一个刚性的U形结构，称为_Socket_。插座是由坚硬的塑料制成的，以承载体重，并手工制作，以确保最好的定制适合遵循树桩的轮廓。然而，75%的截肢者感到不高兴，因为残肢每小时都会改变形状，而他们定制的坚硬的插座却不会。这会导致持续不断的皮肤摩擦，导致疼痛的伤口和溃疡。在截肢者的余生中，每年更换插座4次是目前最好的解决方案，每年在英国、欧盟和美国花费71亿GB的公共资金(不包括创伤护理费用：仅在英国就花费24亿GB(英国-政府/APPG；2019年))。截肢者需要自信和富有成效的生活，尽管他们的病情。因此，帝国理工学院和Unhindr开发了Roliner，让数百万人的生活变得独立。我们通过之前一个成功的Innovate-UK项目证明了这一概念和可制造性，现在我们必须开发一种MVP(最小可行产品)。**内容：**Roliner是一种在安装假肢之前戴在树桩上的袖状设备。**方法：**Roliner使用人工智能(机器学习)来了解树桩每小时/每天的变化并适应它们。罗莱纳读取残肢和插座之间的实时传感器数据，通过应用程序了解截肢者的舒适度偏好，并通过充气/放气罗林纳的微通道无缝而持续地调整配件。**影响：**假体装配是截肢者维持日常活动的最大障碍。截肢者依靠合适的诊所生活。因此，在截肢的第一年内，六分之一的截肢者会失去工作。44.3%的工作年龄截肢者不从事经济活动，造成48.9亿GB的生产力损失(Papworth Trust，2016)。罗林纳的机器学习模仿临床实践，提供健壮的行走体验，将减少生产力损失，减少对医院的依赖，并可能每年为NHS节省13亿GB。在眼窝调整方面，24亿GB的创伤。由于生活质量和活动质量的下降，截肢者迅速失去肌肉。截肢者在35岁时，每天使用假肢行走超过500米的概率为74%；而在60岁时，几乎只能坐轮椅行走的概率仅为34%(来源：Geertzen，Jan et.al；2005-Claimed-walking-distance-of-lower-limb截肢者。-残疾与康复)。凭借首创的无缝AI适配，Roliner可以将行走能力提高3倍，提供健美的行走体验。减少由眼窝引起的伤口可降低感染风险，增加活动能力，从而减少肌肉损失。

**"The Human body changes shape, but prosthetic sockets don't."**Our bodies are approximately 1.5 cm shorter in evenings, compared to mornings. Our feet and hands are larger in evenings. Most of us don't notice these changes because we wear flexible clothes.However, if you lose a leg, you must wear a prosthetic limb attached to your body with a rigid U-shaped structure called a _socket_. Sockets are made of hard plastics to carry the bodyweight and hand-made to ensure the best custom fit by following the stump's contour. However, 75% of amputees are unhappy because the stump changes shape every hour while their custom-made rigid sockets don't. This leads to constant skin rubbing causing painful bleeding wounds and ulcers. Replacing sockets upto 4 times annually for the rest of amputees' lives is currently the best solution costing £7.1billion of public money in the UK, EU and US each year(Excluding woundcare cost: £2.4billion in UK alone(UK-Government/APPG;2019)).Amputees need a confident and productive life despite their conditions. Therefore, Imperial College and Unhindr develop Roliner to make lives of millions independent. We proved the concept and manufacturability with a previous successful Innovate-UK project, and now we must develop an MVP(Minimum-Viable-Product).**What:** Roliner is a sleeve-like device worn on the stump before putting the prosthetic limb on.**How:** Roliner uses Artificial Intelligence(Machine Learning) to understand hourly/daily changes in the stump and adapts to them. Roliner reads real-time sensors' data between the stump and socket, learns amputees' comfort preferences via an app, and seamlessly and continuously adjust the fitting by inflating/deflating Roliner's micro-channels.**Impact:** Prosthesis fitting is the biggest barrier for amputees to maintain their daily activities. Amputees live dependent on fitting clinics. Therefore, within the first year of amputation, 1-in-6 amputees lose their jobs. 44.3% working age amputees are economically inactive, costing £4.89billion in productivity loss(PapworthTrust,2016). Roliner's Machine-Learning mimicking the clinical practice and providing able-bodied-like walking experience will reduce productivity loss, hospital dependency, and potentially save NHS £1.3billion p.a. in socket adjustments, £2.4billion in woundcare.Due to reduced quality-of life and activity, amputees rapidly lose muscles. The probability of an amputee walking with a prosthesis more than 500metres a day is 74% at age 35; while only 34% at age 60, becoming almost wheelchair-bound(Source:Geertzen,Jan et.al;2005-Claimed-walking-distance-of-lower-limb amputees.-Disability-and-rehabilitation). With first-of-its-kind seamless AI-adaptation, Roliner could increase walking performance x3 times, providing able-body-like walking experience. Reducing socket-induced wounds reduces risk of infection, increases mobility, which reduces muscle loss.