AI-powered next-generation imaging biomarkers for dementia

人工智能驱动的下一代痴呆症成像生物标志物

• 批准号: MR/W004097/1
• 负责人: Hui Zhang
• 金额: $ 9.89万
• 依托单位: University College London
• 依托单位国家: 英国
• 项目类别: Research Grant
• 财政年份: 2021
• 资助国家: 英国
• 起止时间: 2021 至 无数据
• 项目状态: 已结题
• 来源: https://gtr.ukri.org/projects?ref=MR%2FW004097%2F1
• 关键词: AI; powered; next; generation; imaging

Dementia is one of the foremost global healthcare challenges of our times, with a vast physical, psychological, social and economic impact. About 50 million people worldwide live with dementia today and this number is expected to triple by 2050. The direct medical and social care costs have been estimated in 2015 to be 1.4% of gross domestic product in high-income countries. Since 2016, dementia has become the leading cause of deaths in the UK. Even today, in the midst of an unprecedented pandemic, Covid-19 is hitting people with dementia the hardest, with dementia being the most common pre-existing condition for people dying of Covid-19. Moreover, patients recovered from Covid-19 are thought to be at significantly increased risk for dementia later in life. Indeed, the need to find a cure for dementia has never been greater.However, till this day, no disease-modifying treatments have been developed. The attrition rate of treatment trials for Alzheimer's disease (AD), the commonest cause of dementia, is high, with 98% failing in Phase III. The reason, at least in part, for the successive, and often costly, trial failures is likely to be that the disease is targeted too late, after significant and irreversible loss of brain cells has occurred. Hence, there is now a consensus and drive towards undertaking trials of therapies much earlier, before the disease becomes irreversible. This calls for the development of more advanced biomarkers that can offer increased sensitivity for identifying individuals at early stages of the disease.University College London (UCL) is an international leader in the development of cutting-edge MRI biomarkers for early diagnosis of dementia. Dr Michele Guerreri, the proposed secondee, is an early-career imaging researcher at UCL who has made significant progress in our quest towards better biomarkers for dementia. This secondment is driven by the aspiration to translate this cutting-edge research into a product to accelerate its real-world impact.AINOSTICS, the hosting organization, is an award-winning imaging AI start-up which specialises in combining cutting-edge MRI and other biomarker data with state-of-the-art artificial intelligence (AI) to detect early changes in AD and other neurodegenerative diseases. They thus represent a perfect partner to deliver this project.This project will boost the skills of Dr Guerreri which will be working in a cutting-edge AI environment, learning the translation process from a research idea into a commercial product. The Secondee will have the opportunity to experience the commercial sector for the first time, learning the development and commercialisation process of his academic research. The hosting organisation will benefit from the Secondee's expertise in the development of state-of-the-art imaging biomarkers for dementia.In the long term, the project will lay the groundwork for a close collaboration between UCL and AINOSTICS. The expertise gained by Dr Guerreri will allow him to develop similar relationships with other players in the vibrant UK imaging AI sector, closing the traditional gap that has been preventing or slowing the knowledge exchange between biomedical industry and academia.

痴呆症是我们这个时代最重要的全球医疗保健挑战之一，具有巨大的身体，心理，社会和经济影响。目前，全球约有5000万人患有痴呆症，预计到2050年这一数字将增加两倍。据估计，2015年高收入国家的直接医疗和社会保健费用占国内生产总值的1.4%。自2016年以来，痴呆症已成为英国死亡的主要原因。即使在今天，在前所未有的大流行中，Covid-19对痴呆症患者的打击最为严重，痴呆症是死于Covid-19的人最常见的既往疾病。此外，从新冠肺炎中康复的患者被认为在以后的生活中患痴呆症的风险显著增加。事实上，寻找治疗痴呆症的方法的需求从未如此强烈。然而，直到今天，还没有开发出改善疾病的治疗方法。阿尔茨海默病（AD）是痴呆症最常见的原因，其治疗试验的损耗率很高，98%的III期试验失败。连续的、往往代价高昂的试验失败的原因，至少部分是因为这种疾病的靶向治疗太晚，在脑细胞发生重大和不可逆转的损失之后。因此，现在有一个共识，并推动在疾病变得不可逆转之前更早地进行治疗试验。这就需要开发更先进的生物标志物，以提高在疾病早期阶段识别个体的灵敏度。伦敦大学学院（UCL）在开发用于痴呆症早期诊断的尖端MRI生物标志物方面处于国际领先地位。Michele Guerreri博士是UCL的早期职业成像研究员，他在我们寻求更好的痴呆症生物标志物方面取得了重大进展。AINOSTICS是一家屡获殊荣的人工智能成像初创公司，专注于将最先进的MRI和其他生物标志物数据与最先进的人工智能（AI）相结合，以检测AD和其他神经退行性疾病的早期变化。该项目将提高Guerreri博士的技能，他将在尖端的人工智能环境中工作，学习将研究想法转化为商业产品的过程。被派遣者将有机会第一次体验商业部门，学习其学术研究的发展和商业化过程。主办机构将受益于借调人员在开发最先进的痴呆症成像生物标志物方面的专业知识。从长远来看，该项目将为UCL和AINOSTICS之间的密切合作奠定基础。Guerreri博士获得的专业知识将使他能够与充满活力的英国成像人工智能领域的其他参与者建立类似的关系，缩小阻碍或减缓生物医学行业与学术界之间知识交流的传统差距。

项目成果

Deep learning framework for accelerating revised-NODDI parameter estimation with tensor-valued diffusion encoding

使用张量值扩散编码加速修订版 NODDI 参数估计的深度学习框架

• 发表时间: 2022
• 作者: Guerreri M.

Resolving quantitative MRI model degeneracy with machine learning via training data distribution design

通过训练数据分布设计利用机器学习解决定量 MRI 模型简并性

• 发表时间: 2023
• 作者: Guerreri M

Revised-NODDI with conventional dMRI data enabled by deep learning

深度学习支持的传统 dMRI 数据的修订版 NODDI

• 发表时间: 2022
• 作者: Guerreri M.

Can machine learning resolve model degeneracy in tissue microstructure estimation?

机器学习能否解决组织微观结构估计中的模型简并性？

• 发表时间: 2023
• 作者: Michele Guerreri

Information Processing in Medical Imaging - 28th International Conference, IPMI 2023, San Carlos de Bariloche, Argentina, June 18-23, 2023, Proceedings

医学影像信息处理 - 第 28 届国际会议，IPMI 2023，阿根廷圣卡洛斯-德巴里洛切，2023 年 6 月 18-23 日，会议记录

• DOI: 10.1007/978-3-031-34048-2_1
• 发表时间: 2023
• 作者: Guerreri M