Improving an EEG-based neurodiagnostic software platform to detect Alzheimer's Disease in MCI patients

改进基于脑电图的神经诊断软件平台来检测 MCI 患者的阿尔茨海默病

• 批准号: 10546255
• 负责人: Richard J. Caselli
• 金额: $ 29.98万
• 依托单位: SPARK NEURO INC.
• 依托单位国家: 美国
• 财政年份: 2022
• 资助国家: 美国
• 起止时间: 2022-09-15 至 2023-08-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/10546255
• 关键词: Accounting; Address; Affect; Algorithms; Alzheimer disease screening; Alzheimer' s Disease; Alzheimer' s disease care; Alzheimer' s disease diagnosis; Alzheimer’s disease biomarker; Behavioral Symptoms; Biological Markers; Clinic; Clinical; Cognitive; Collaborations; Computer software; Country; Data; Data Pooling; Dementia; Detection; Diagnosis; Disease; Disease Progression; Early Diagnosis; Elderly; Electroencephalography; Ensure; Feedback; Health; Health Care Costs; Human; Impaired cognition; Machine Learning; Measures; Medical; Methods; Mission; Modeling; Nerve Degeneration; Neurologic; Neurologist; Outcome Measure; Pathologic; Pathway interactions; Patient Care; Patients; Performance; Persons; Phase; Physicians; Process; Psyche structure; Reporting; Research; Scanning; Sensitivity and Specificity; Specificity; Standardization; Symptoms; System; Techniques; Technology; Testing; Time; Tissues; Training; United States; Work; aging population; base; clinical care; cost; cost effective; data acquisition; diagnostic tool; experience; human old age (65+); improved; insight; mild cognitive impairment; minimally invasive; novel; novel marker; preservation; satisfaction; screening program; success; tool; usability; user-friendly

PROJECT SUMMARY Alzheimer’s disease (AD) is a progressive, neurodegenerative condition and the most common cause of dementia. In the United States, an estimated 6.2 million people over the age of 65 are living with AD, 72% of whom are over 75 years old. Given the country’s aging population, this number is expected to more than triple by 2050, costing the United States an annual $600 billion in associated healthcare costs. Early diagnosis is crucial to AD treatment because it allows clinicians more time to find and initiate treatment pathways, which decreases disease progression and preserves mental capacity. New research suggests that biomarkers can help diagnose AD years before symptoms appear. Despite recent technological advancements, many tools and technologies that measure biomarkers are invasive, expensive, and not sensitive or specific enough, particularly when detecting the disease at earlier stages, limiting their usability. When combined with advanced machine learning techniques, electroencephalography (EEG) has been shown to address many of the existing issues related to AD biomarkers. At SPARK Neuro, we aim to unlock the full potential of EEG through a novel software platform. Combining EEG with the capabilities of machine learning, our model better assesses cognitive health and neurodegeneration, aiding the diagnosis of AD. SPARK’s neuroanalytic platform will be a standardized, objective, non-invasive, cost-effective diagnostic tool capable of highly sensitive and specific detection of cognitive impairment across the entire disease continuum. Our platform would vastly expand AD screening initiatives and provide neurological insights to aid in the diagnosis and tracking of disease progression. During the proposed Phase I research, we will work in collaboration with Mayo Clinic to extend our current algorithm to assess and differentiate patients in the earlier, mild cognitive impairment stage of the disease, and provide highly useful and usable reports to clinicians. First, we will optimize the algorithm by incorporating EEG data collected from Mayo Clinic patients. Next, we will focus on improving the user experience of both EEG data acquisition and clinical reporting. We will enhance end-user satisfaction and optimize the technology to fit within current clinical workflows. Participating Mayo Clinic EEG technicians will provide feedback. Once optimized, SPARK’s approach will constitute the first in-office EEG-based neurodiagnostic tool specifically for diagnosing and tracking AD. Our non-invasive solution has the potential to accelerate AD screening programs, detect pathological AD at earlier stages, and provide individualized disease progression insights.

项目概要 阿尔茨海默病 (AD) 是一种进行性神经退行性疾病，也是导致该病的最常见原因 在美国，估计有 620 万名 65 岁以上的人患有 AD，其中 72% 鉴于该国人口老龄化，这一数字预计将增加两倍以上。 到 2050 年，美国每年因早期诊断而花费的相关医疗费用将达到 6000 亿美元。 对于 AD 治疗至关重要，因为它可以让我们有更多时间寻找和启动治疗途径，从而 新的研究表明，生物标志物可以减缓疾病进展并保持心智能力。 尽管最近技术进步，但仍有许多工具和工具可以在症状出现之前帮助诊断 AD。 测量生物标志物的技术是侵入性的、昂贵的、并且不够灵敏或特异性，特别是 在早期阶段检测疾病时，限制了它们的可用性。 当与先进的机器学习技术相结合时，脑电图（EEG）已被证明 为了解决与 AD 生物标志物相关的许多现有问题，在 SPARK Neuro，我们的目标是解锁完整的解决方案。 通过新颖的软件平台将脑电图与机器学习功能相结合， 我们的模型可以更好地评估认知健康和神经退行性疾病，帮助诊断 SPARK。 神经分析平台将是一个标准化、客观、非侵入性、具有成本效益的诊断工具，能够 我们的平台对整个疾病连续体的认知障碍进行高度敏感和特异性的检测。 将极大地扩大 AD 筛查计划并提供神经学见解以帮助诊断和跟踪 疾病进展。 在拟议的第一阶段研究期间，我们将与梅奥诊所合作，以扩展我们目前的 用于评估和区分处于疾病早期、轻度认知障碍阶段的患者的算法，以及 首先，我们将通过合并脑电图来优化算法。 接下来，我们将重点改善这两种脑电图数据的用户体验。 我们将提高最终用户的满意度并优化技术以适应。 当前的临床工作流程一旦优化，参与梅奥诊所脑电图技术人员将提供反馈。 SPARK 的方法将构成第一个基于办公室脑电图的神经诊断工具，专门用于诊断 我们的非侵入性解决方案有可能加速 AD 筛查计划、检测。 早期阶段的病理性 AD，并提供个性化的疾病进展见解。