RATify - Radiology AI Assurance Testing platform

RATify - 放射学人工智能保证测试平台

• 批准号: 10074053
• 金额: $ 6.32万
• 依托单位: ENVISIONIT DEEP AI LTD
• 依托单位国家: 英国
• 项目类别: Grant for R&D
• 财政年份: 2023
• 资助国家: 英国
• 起止时间: 2023 至 无数据
• 项目状态: 已结题
• 来源: https://gtr.ukri.org/projects?ref=10074053
• 关键词: RATify; Radiology; AI; Assurance; Testing

AI-based healthcare applications (AI) could potentially have a significant impact on modernising health care systems globally, with radiology seen as one of the most promising fields where clinical AI offers a potential solution to the increasing demand for imaging tests and the ongoing workforce crisis (BMC HEalth SErvices Research 2021).However, as AI adoption expands into radiology, there is growing concern that AI algorithms need to undergo quality assurance reviews, similar to imaging systems, to ensure they are safe and effective for patients, performing as intended and producing accurate data.Whilst internal validation is required for regulatory compliance (MHRA) there are currently no standard industry pathways to validate AI medical imaging solutions once deployed. However the regulation landscape for AI med-tech is changing rapidly, particularly in the UK with the NICE ESF digital health technologies 2022 update requiring demonstration of value in the UK health and social care system.Envisionit's new AI validation platform, RATify, is the first to perform both retrospective and prospective AI studies on classification, object detection and segmentation AI models, as specified in the NICE 2022 updates.The vendor agnostic AI Validation platform provides holistic AI assurance for organisations when selecting, deploying and monitoring AI solutions in their environment and a seamless workflow for radiologists for real-time audit and monitoring. The tool will also be relevant to wider stakeholders in the AI adoption life cycle including AI developers, service providers and researchers.The successful commercialisation of RATify will underpin our global scaling ambitions, whilst also boosting confidence and implementation of AI technologies into radiology and solving a global need for clinical governance, specifically relevant for the NHS.

基于人工智能的医疗保健应用（AI）可能对全球医疗保健系统的现代化产生重大影响，放射学被视为最有前途的领域之一，临床AI为日益增长的成像测试需求和持续的劳动力危机提供了潜在的解决方案（BMC HEALTH SERVICE研究2021）。然而，随着人工智能的采用扩展到放射学，人们越来越担心人工智能算法需要接受质量保证审查，与成像系统类似，以确保它们对患者安全有效，按预期运行并产生准确的数据。虽然监管合规性（MHRA）需要内部验证，但目前还没有标准的行业途径来验证AI医疗成像解决方案。然而，人工智能医疗技术的监管格局正在迅速变化，特别是在英国，NICE ESF数字健康技术2022更新要求在英国健康和社会保健系统中展示价值。Envisionit的新人工智能验证平台RATify是第一个对分类，对象检测和分割人工智能模型进行回顾性和前瞻性人工智能研究的平台，与供应商无关的AI验证平台为组织在其环境中选择、部署和监控AI解决方案时提供全面的AI保证，并为放射科医生提供无缝的工作流程，以进行实时审计和监控。该工具还将与人工智能采用生命周期中更广泛的利益相关者相关，包括人工智能开发人员，服务提供商和研究人员。RATify的成功商业化将支持我们的全球扩展雄心，同时也将增强人们对人工智能技术在放射学中的信心和实施，并解决全球对临床治理的需求，特别是与NHS相关。