Intelligent DEPression Tracking (I-DEPT) - A Novel, Longitudinal and Multi-Modal Machine Learning Framework for Quantifying Depression Symptoms

智能抑郁追踪 (I-DEPT) - 一种新颖的、纵向的、多模态的机器学习框架，用于量化抑郁症状

• 批准号: 10050419
• 金额: $ 29.99万
• 依托单位: THYMIA LIMITED
• 依托单位国家: 英国
• 项目类别: Collaborative 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: 未结题
• 来源: https://gtr.ukri.org/projects?ref=10050419
• 关键词: Intelligent; DEPression; Tracking; DEPT; Novel

Depression is a massive, growing economic and societal problem; it is a leading cause of disability and suicides, costing the UK economy \>£56bn annually in lost productivity. There is an urgent need to relieve this burden on the economy and the NHS.The healthcare sector, however, needs better tools to tackle the problem of depression efficiently. Its heterogeneous clinical profile means that patients can have several unique combinations of depressive symptoms. Currently, identifying the right diagnosis and treatment in the UK may take many years, with some studies finding rates of untreated depression as high as 77%. There is an urgent need for a tool that can help clinicians objectively measure individual depression _symptoms and symptom clusters_ - just as physical illness ones (e.g. blood test markers).Artificial intelligence (AI) tools have been proposed to automatically detect depression (but not symptoms like mood, fatigue or anhedonia). They also face barriers to commercial implementation given their limited validity and generalisability. This is due to their focusing on small samples, a single time point and/or a single behaviour modality (e.g. voice, which reduces accuracy and sensitivity).**The aim of this project is to bridge this gap by developing an innovative healthcare solution in the form of a next generation, reliable AI-powered depression screening tool for clinicians ("I-DEPT").**To this end, we will collect several minutes' worth of gamified activities twice a week, over 3 months, from 550 depressed individuals and 550 controls via an online recruitment platform (Prolific; e.g. speech elicitation, working memory etc.).Ultimately, our end-to-end clinician support tool will save clinicians significant admin time and costs, doubling patient throughput and revenue, halving waiting times, all whilst improving clinical outcomes. The project will take us from proof-of-concept to a fully commercialisable product with strong, demonstrable return on investment.

抑郁症是一个巨大的，日益严重的经济和社会问题;它是残疾和自杀的主要原因，每年造成英国经济损失超过560亿英镑。迫切需要减轻经济和NHS的负担。然而，医疗保健部门需要更好的工具来有效地解决抑郁症问题。其异质性的临床特征意味着患者可以有几种独特的抑郁症状组合。目前，在英国确定正确的诊断和治疗可能需要多年时间，一些研究发现未经治疗的抑郁症比例高达77%。人们迫切需要一种工具，可以帮助临床医生客观地测量个体抑郁症-症状和症状群-就像身体疾病一样（例如血液测试标记物）。人工智能（AI）工具已经被提出来自动检测抑郁症（但不是情绪，疲劳或快感缺乏等症状）。由于其有效性和普遍性有限，它们在商业实施方面也面临障碍。这是因为他们专注于小样本，单一时间点和/或单一行为模式（例如语音，这降低了准确性和灵敏度）。该项目的目的是通过开发一种创新的医疗保健解决方案来弥合这一差距，该解决方案的形式是为临床医生提供下一代可靠的人工智能抑郁症筛查工具（“I-DEPT”）。为此，我们将通过在线招募平台（Prolific;例如，言语诱导，工作记忆等），每周两次，在3个月内从550名抑郁症患者和550名对照组中收集几分钟的游戏化活动。最终，我们的端到端临床医生支持工具将为临床医生节省大量的管理时间和成本，使患者吞吐量和收入翻一番，将等待时间减半，同时改善临床结果。该项目将把我们从概念验证带到完全商业化的产品，并带来强劲的投资回报。