Development of an AI-empowered device that utilizes multimodal data-visualization to aid in the diagnosis, and treatment, of OUD

开发人工智能设备，利用多模式数据可视化来帮助 OUD 的诊断和治疗

• 批准号: 10683816
• 负责人: DAVID OAKLEY
• 金额: $ 31.91万
• 依托单位: WAVI COMPANY
• 依托单位国家: 美国
• 财政年份: 2023
• 资助国家: 美国
• 起止时间: 2023-09-01 至 2024-08-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/10683816
• 关键词: Address; Admission activity; Algorithms; Brain; Classification; Clinic; Clinical; Clinical Trials; Cognitive; Computer software; Data; Development; Devices; Diagnosis; Disease Management; Electroencephalography; Intake; Intervention; Marketing; Measurement; Measures; Monitor; Patients; Phase; Phenotype; Population; Reaction Time; Rehabilitation Outcome; Rehabilitation therapy; Relapse; Scanning; Sensitivity and Specificity; Testing; Training; addiction; artificial intelligence algorithm; cognitive change; cognitive testing; commercialization; cost; cost effective; data visualization; effectiveness evaluation; empowerment; experience; follow-up; heart rate variability; improved outcome; multimodal data; multimodality; opioid use disorder; outcome prediction; programs; recidivism; response; social; tool; treatment response; user-friendly

Abstract Assessing the effectiveness of opioid use disorder (OUD), where high relapse rates create financial and social tolls, is a pressing clinical problem in need of better measurement tools. The ability to identify cognitive changes should improve outcomes but current intake into rehabilitation programs doesn’t typically include cognitive tests. Simple, quick, cost-effective, and objective measures are needed. This is the problem this proposal seeks to address. This proposal utilizes the experience of 8 different rehab clinics which serve 150 patients weekly, and WAVi, a commercialized brain-assessment platform that combines EEG evoked responses (ERP) with 5 other tests also sensitive to addiction (heart rate variability, physical reaction times, MoCA, Trail Making, and Flanker). This user-friendly platform focuses on minimizing testing times and cost while maximizing information. For this fast-track application, we will have the following milestones: - Collect more OUD data from different clinics to refine existing clustering algorithm and increase sensitivities and specificities so that we have a robust archetype for OUD vs healthy patients - Collect follow-up OUD data and correlate follow-up scans with successful outcomes of rehabilitation treatment and therefore identify those addicts whose cognitive state requires modified treatment approaches, with the aim of decreasing relapse rates and recidivism rates. - Develop a scalable multimodal product, including EEG with ERP, for rehabilitation facilities that is readily accessible to clinicians and create a dynamic data asset to help longitudinally predict outcomes.

摘要 评估阿片类药物使用障碍（OUD）的有效性，其中高复发率造成 财务和社会费用，是一个迫切的临床问题，需要更好的测量工具。 识别认知变化的能力应该会改善结果，但目前的摄入量 康复项目通常不包括认知测试简单、快速、经济高效， 需要客观的措施。这就是这项建议试图解决的问题。 这项建议利用了8个不同康复诊所的经验，为150名病人提供服务 每周，和WAVi，一个商业化的大脑评估平台，结合脑电图诱发 反应（ERP）与其他5个测试也敏感成瘾（心率变异性，身体 反应时间，莫卡，Trail Making和Flanker）。这个用户友好的平台专注于 最小化测试时间和成本，同时最大化信息。 对于这个快速通道应用程序，我们将有以下里程碑： - 从不同诊所收集更多OUD数据，以完善现有聚类算法 并增加敏感性和特异性，使我们有一个强大的原型OUD 与健康患者相比 - 收集随访OUD数据并将随访扫描与成功结局相关联 从而识别出那些认知状态 需要改进治疗方法，以降低复发率， 累犯率。 - 开发可扩展的多模式产品，包括EEG和ERP，用于康复 临床医生可以随时访问的设施，并创建动态数据资产， 纵向预测结果。