Industrial research project of new real-time guiding system for surgeons' operations using Artificial Intelligence

这个工业研究项目旨在验证我们现有的人工智能引导系统原型，旨在指导外科医生在更短的时间内进行髓内钉（IM）手术，减少辐射，减少对患者和医务人员的风险。几十年来，IM的手术技术基本保持不变。IM手术用于帮助骨折的焊接，每年在全球范围内进行超过160万次。不幸的是，按照目前的方法，一个20分钟的简单手术可能需要长达2个小时的时间，而且由于难以确定钉子的位置，需要大量的x光片。这些挑战可能意味着手术团队和患者暴露于辐射的增加，以及麻醉和失血增加对患者的额外风险。NHS每年仅花在髓内钉手术上的费用就超过5亿英镑（每位患者约1万英镑），我们可以减少手术时间、辐射、风险和成本，同时提高患者满意度，减少取消手术和住院时间，并清除等待名单。在与来自英国和西班牙的135名不同级别的整形外科医生交谈后，93%的人认为实施这种手术并不总是那么容易，另外97%的人渴望看到市场上的新解决方案来指导他们。该原型使用了一系列人工智能算法，包括增强现实、跟踪算法和计算机视觉，在混合现实耳机上实现，提供了一种无需手动的解决方案，能够向外科医生展示3D实时反馈，而不是依赖于2D静态x射线。

This industrial research project aims to validate our existing Artificial Intelligence guiding system prototype aimed at guiding surgeons during Intramedullary Nailing (IM) surgeries in less time, with less radiation, generating fewer risks to both patients and medical personnel.Surgical techniques for IM have remained largely unchanged for decades. IM surgery is used to assist in the welding of broken bones and is performed over 1.6 million times per year worldwide.Unfortunately, what should be a straightforward operation taking 20 minutes, can take as long as 2 hours as per the guesswork involved in the current methodology and require a large number of X-rays due to the difficulty in identifying where the nails need to be placed. These challenges can mean increased exposure to radiation for the surgical team and patients and additional risks for patients from increased anaesthetics and blood loss.With over £500 million spent by the NHS only in intramedullary nailing surgeries every year (around £10,000 per patient), we can reduce operating time, radiation, risks and costs while increasing patient satisfaction, reducing cancelled operations and length of stay and clearing waiting lists.After talking to 135 orthopaedic surgeons, at different levels, from the UK and Spain, 93% of them agreed it is not always easy to perform this surgery and a further 97% are keen to see a new solution in the market to guide them.This prototype uses a series of Artificial Intelligence algorithms including Augmented Reality, tracking algorithms and computer vision, implemented on a Mixed Reality headset to provide a hands-free solution capable of showing surgeons in 3D real-time feedback of where to operate rather than relying on 2D static X-Ray.