AI-PROGNOSIS: Artificial intelligence-based Parkinson’s disease risk assessment and prognosis

AI-PROGNOSIS：基于人工智能的帕金森病风险评估和预后

• 批准号: 10080508
• 金额: $ 17.13万
• 依托单位: UNIVERSITY OF OXFORD
• 依托单位国家: 英国
• 项目类别: EU-Funded
• 财政年份: 2023
• 资助国家: 英国
• 起止时间: 2023 至 无数据
• 项目状态: 未结题
• 来源: https://gtr.ukri.org/projects?ref=10080508
• 关键词: AI; PROGNOSIS; Artificial; intelligence; based

Parkinson’s disease (PD) is the most common neurodegenerative movement disorder, with a multifactorial aetiology, heterogeneous manifestation of motor and non-motor symptoms, and no cure. PD is often missed or misdiagnosed, as early symptoms are subtle and common with other diseases, allowing for considerable damage to occur before treatment. Moreover, selecting the optimal medication regimen is usually a lengthy, “trial and error” process, leading to critical, costly non-adherence. Following a trustworthy and inclusive approach to AI development and based on multidisciplinary expertise and broad stakeholder engagement, AI-PROGNOSIS aims to advance PD diagnosis and care by: 1) developing novel, predictive AI models for personalised PD risk assessment and prognosis (in terms of time to higher disability transition and response to medication) based on multi-source patient records and databases, including in-depth health, phenotypic and genetic data, 2) implementing a system of biomarkers informing the AI models by tracking key risk/ progression markers in daily living, and ultimately 3) translating the models and digital biomarkers into a validated, privacy-aware AIdriven toolkit, supporting healthcare professionals (HCPs) in disease screening, monitoring and treatment optimization via quantitative, explainable evidence, and empowering individuals with/without PD with tailored insights for informed health management.

帕金森病(PD)是最常见的神经退行性运动障碍，具有多因素的病因，运动和非运动症状的表现不一，无法治愈。帕金森病经常被漏诊或误诊，因为早期症状微妙，与其他疾病一样常见，因此在治疗前可能会发生相当大的损害。此外，选择最佳用药方案通常是一个漫长的“反复试验”过程，导致关键的、代价高昂的不依从性。遵循值得信赖和包容的人工智能发展方法，基于多学科专业知识和广泛的利益相关者参与，AI-预后旨在通过以下方式推进PD诊断和护理：1)基于多源患者记录和数据库，包括深入的健康、表型和遗传数据，开发用于个性化PD风险评估和预后(关于更高残疾过渡时间和药物反应)的新颖、预测性AI模型；2)实施生物标记物系统，通过跟踪日常生活中的关键风险/进展标记来通知AI模型；以及最终3)将模型和数字生物标记物转化为有效的、隐私感知的AI驱动工具包，通过量化、可解释的证据支持医疗保健专业人员(HCP)进行疾病筛查、监测和治疗优化，并使患有/不患有帕金森病的个人能够获得针对知情健康管理的量身定制的见解。