CESCAIL - Capsule Endoscopy delivery at SCale through enhanced AI anaLysis

我们的目标是通过预防、早期检测和诊断的集成解决方案，帮助根除肠癌并改进各种胃肠道疾病的治疗。Corporation Health International(CHI)的主要产品是远程内窥镜管理服务：胶囊内窥镜(CE)可以通过可吞咽的相机对肠道进行可视化，并提供全面评估以确定肠癌风险、诊断GI症状，并准确决定最佳后续治疗(如果有的话)。这项服务可以在任何情况下提供，从门诊到自我管理。目前，CHI是NHS苏格兰与CE制造商美敦力合作的唯一服务提供商，也是唯一家在整个欧洲提供这种受管服务的商业公司。英国NHS正在评估他们部署CE的方法，这可能会为CHI提供一个广阔的市场准入。视频分析是一个资源密集型过程，依赖于专家护士和医生，他们在两个阶段的过程中检查多达40万张图像以发现可疑发现，需要长达2小时的时间才能获得高质量的报告。人工智能(AI)承诺可以在提高整体准确性的同时提高这一过程的效率。CHI在为基于神经网络的AI解决方案创建机器学习(ML)数据以及在我们的运营环境中使用该AI的组织灵活性方面具有独特的地位。自2019年1月以来，CHI一直在与多家公司、学者和NHS组织合作，参与由英国发起的创新项目AID-GI，以开发这样的人工智能系统。AID-GI将于2021年2月28日结束，提供一个原型，准备与标准操作一起使用。因此，我们的项目希望利用原型来进行所有必要的临床试验以进行认证，强化软件以达到医疗器械IIa类状态，并进一步提高AI系统的广度和准确性。然后，结果应该是AI主导的过程，将平均视频分析时间减少到30分钟以下。反过来，这又减少了资源需求，使真正大规模的服务成为可能。为抓住英国市场机遇而招聘需要100周的工作，现在可以在10周内完成。

Our aim is to help eradicate bowel cancer and improve treatment for a wide range of gastrointestinal diseases through integrated solutions for prevention, early detection, and diagnosis.CorporateHealth International's (CHI) primary product is a tele-endoscopy managed service: Capsule Endoscopy (CE) to visualise the bowel with a swallowable camera and provide a full evaluation to determine bowel cancer risk, diagnose GI symptoms, and precisely decide on optimal follow-up -- if any. This service can be delivered in any setting, from outpatient to self-administration.Currently, CHI is the single service provider for NHS Scotland in a partnership with the CE manufacturer Medtronic, and only commercial company to offer this managed service across Europe. NHS England is evaluating their approach to deploy CE, potentially offering CHI a broad market entry.Video analysis is a resource-intensive process, relying on expert nurses and doctors that inspect up to 400,000 images for suspicious findings in a 2-stage process taking up to 2 hours for a high-quality report.Artificial Intelligence (AI) holds the promise that this process can be made much more efficient while increasing overall accuracy.CHI is in a unique position to create Machine Learning (ML) data for a Neural Network-based AI solution as well as to have the organizational flexibility to make use of that AI within our operational environment.Since January 2019, CHI has been partnering with multiple companies, academics, and NHS organizations in the Innovate UK-sponsored project AID-GI to develop such an AI system.AID-GI concludes February 28, 2021 with a prototype, ready to be used alongside standard operations.Our project is therefore looking to leverage the prototype to conduct all necessary clinical trials for certification, harden the software to achieve medical device Class IIa status and further improve the breadth and accuracy of the AI system.The outcome should then be an AI-led process that reduces the average video analysis time to under 30 minutes. In turn that reduces resource requirements to make a truly large-scale service possible. What would take 100 weeks to recruit to address the UK market opportunity, can then be done in 10 weeks.