Artificial General Intelligence in support of General Practice Healthcare

支持全科医疗保健的通用人工智能

• 批准号: 133389
• 金额: $ 6.69万
• 依托单位: API COGNITION LIMITED
• 依托单位国家: 英国
• 项目类别: Feasibility Studies
• 财政年份: 2018
• 资助国家: 英国
• 起止时间: 2018 至 无数据
• 项目状态: 已结题
• 来源: https://gtr.ukri.org/projects?ref=133389
• 关键词: Artificial; General; Intelligence; support; Practice

"We face a crisis in the provision of primary health care. An increasing number of online technology solutions are available to help support General Practitioners (GPs). However, they have either failed to gain widespread support amongst GPs, or amongst the patients which they are intended to serve. At one end of the spectrum there are services which offer diagnostic services, employing scripted questions generated by decision tree algorithms, that fail to exhibit common sense or engage with the patient in a humanistic fashion. At the other end there are support systems that are backed by real doctors, but fail to offer patients a response in real-time.Our approach is to employ artificial intelligence techniques with some diagnostic capacity, in a fashion that resembles the manner in which doctors usually interact with patients, something far closer to the ""bed-side-manner"" that we are used to. We are able to do this through the use of biologically inspired Artificial General Intelligence (AGI) developed over a period of more than ten years. One of the key differences is that the artificial agent that embodies these diagnostic skills, also has an awareness of the consultation process itself, and not only decides what to ask, but also decides when it should listen.The deliverable from this project is the demonstration of virtual doctor, which enters into a consultation with a virtual patient in a stand-alone software simulation. A proof of principle in this project will allow us to demonstrate a patient-friendly real-time doctor's assistant. We are working closely with one existing service provider, and employ staff with both extensive and proven experience in primary care and in artificial intelligence development."

“我们在提供初级卫生保健方面面临危机。越来越多的在线技术解决方案可用于帮助支持全科医生（GP）。然而，他们要么没有得到广泛的支持，在全科医生，或在病人之间，他们打算服务。在频谱的一端，有一些服务提供诊断服务，采用决策树算法生成的脚本问题，未能表现出常识或以人文方式与患者互动。在另一端，支持系统由真实的医生提供支持，但无法为患者提供实时响应。我们的方法是采用具有一定诊断能力的人工智能技术，其方式类似于医生通常与患者互动的方式，更接近于我们习惯的“床边方式“。我们能够通过使用生物启发的人工通用智能（AGI）来实现这一目标，该智能已经开发了十多年。一个关键的区别是，体现这些诊断技能的人工代理也知道咨询过程本身，不仅决定询问什么，还决定何时应该倾听。该项目的可交付成果是虚拟医生的演示，它在独立软件模拟中与虚拟患者进行咨询。在这个项目中的原则证明将使我们能够展示一个病人友好的实时医生助理。我们正在与一家现有的服务提供商密切合作，并聘请在初级保健和人工智能开发方面拥有丰富经验的员工。"