Identification of cognitive decline and dementia: Prediction by everyday driving behaviors and physiological responses

识别认知能力下降和痴呆：通过日常驾驶行为和生理反应进行预测

• 批准号: 10670248
• 负责人: BRUNO GIORDANI
• 金额: $ 105.28万
• 依托单位: UNIVERSITY OF MICHIGAN AT ANN ARBOR
• 依托单位国家: 美国
• 财政年份: 2020
• 资助国家: 美国
• 起止时间: 2020-09-15 至 2025-05-31
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10670248
• 关键词: Age; Alzheimer' s Disease; Alzheimer' s disease related dementia; Alzheimer' s disease risk; Amyloid; Arousal; Automobile Driving; Behavior; Behavior assessment; Behavioral; Brain; Brain Diseases; Breathing; Categories; Characteristics; Classification; Clinical; Cognitive; Collaborations; Collection; Complex; Consensus; Data; Dementia; Detection; Development; Diagnostic; Early Diagnosis; Elderly; Enrollment; Evaluation; Expert Systems; Eye; Funding; Galvanic Skin Response; Goals; Health Personnel; Heart Rate; Impaired cognition; Individual; Language Development; Longitudinal cohort; Machine Learning; Measurement; Measures; Methods; Michigan; Modeling; Monitor; Motor; Neurologic; Neuropsychology; Participant; Pattern; Performance; Personal Satisfaction; Physiological; Population; Positron-Emission Tomography; Ramp; Research; Research Institute; Safety; Sampling; Short-Term Memory; Skin Temperature; Technology; Testing; Time; clinical diagnosis; cognitive performance; cost effective; design; diagnostic accuracy; diagnostic platform; driving behavior; executive function; follow-up; hazard; implementation intervention; learning strategy; machine learning method; meetings; mild cognitive impairment; novel strategies; older driver; programs; psychological aspect of aging; rate of change; recruit; response; smart watch; social; visual tracking; young adult; β-amyloid burden

As the population continues to age and rates of late-life cognitive impairment rise, early detection of cognitive impairment is increasingly important for the timely implementation of interventions and safety initiatives. This may be particularly important in individuals found to have high brain amyloid burden, putting them at particular risk for Alzheimer’s disease and related disorders (ADRD) of the brain. Performance changes in challenging, complex, high-stakes daily activities, such as driving, and accompanying physiological responses may together provide an inexpensive avenue for early detection. This may serve the dual purpose of alerting individuals or health care providers to early cognitive impairment, as well as to potential safety issues. Sophisticated in-car technology that is increasingly becoming standard in new vehicles may provide the means to unobtrusively capture sensitive information about naturalistic driving behaviors and potentially assist with early detection of cognitive impairment. The proposed study will apply a novel approach to unobtrusively monitor older drivers in (a) naturalistic, (b) fixed course, and (c) simulator driving situations. Machine learning approaches will be used to select key features of driving behaviors and physiological measures of arousal in all driving scenarios and eye tracking measures from fixed and simulator drives to predict drivers’ clinical diagnosis: young adult drivers, healthy older drivers with and without high amyloid burden, and drivers with mild cognitive impairment with evident amyloid burden. The participants will be followed longitudinally in the Michigan Alzheimer’s Disease Research Center (MADRC) with annual cognitive and neurological evaluations, as well as repeat driving and physiological testing at two years from baseline. Understanding and identifying changes in driving behaviors and how these predict who will develop clinically identifiable cognitive impairment will lead to the development of a model for early detection of cognitive decline and ADRD.

随着人口的不断老龄化和老年认知障碍率的上升， 损伤对于及时实施干预措施和安全举措越来越重要。这 可能在大脑淀粉样蛋白负荷高的个体中特别重要， 阿尔茨海默病和相关疾病（ADRD）的风险。在挑战中的表现变化， 复杂的、高风险的日常活动，如驾驶，以及伴随的生理反应， 为早期检测提供了一个廉价的途径。这可以起到提醒个人或 医疗保健提供者对早期认知障碍，以及潜在的安全问题。先进的车内 越来越成为新车标准的技术可以提供不引人注目的方式， 捕捉有关自然驾驶行为的敏感信息，并可能有助于早期检测 认知障碍这项拟议的研究将采用一种新的方法来不引人注目地监测老年司机， (a)自然的，（B）固定路线，和（c）模拟器驾驶情况。将使用机器学习方法 在所有驾驶场景中选择驾驶行为的关键特征和唤醒的生理测量， 从固定和模拟驾驶中进行眼动跟踪测量以预测驾驶员的临床诊断：年轻的成年驾驶员， 有和没有高淀粉样蛋白负担的健康老年驾驶员，以及有轻度认知障碍的驾驶员， 明显的淀粉样蛋白负荷。参与者将在密歇根州阿尔茨海默氏症的纵向跟踪 研究中心（MADRC）进行年度认知和神经评估，以及重复驾驶和 从基线起两年进行生理测试。理解和识别驾驶行为的变化 以及如何预测哪些人会出现临床上可识别的认知障碍， 早期检测认知能力下降和ADRD的模型。