ROBAS: A Multimodal Sensor System for Remote Assessment of Oral Health Behaviors

 描述（由申请人提供）：牙科疾病是美国最常见的慢性疾病之一，对个人和社会造成巨大损失。大量研究将牙科疾病与不良口腔卫生行为（例如刷牙和使用牙线不足）联系起来，但未能充分解释这些假定的因果关系是如何产生或介导的。传统的回顾性自我报告容易受到许多偏见的影响，并且缺乏严格检验因果路径假设所需的粒度和科学严谨性。为了解决口腔行为测量的棘手挑战，我们组建了一个由科学家和工程师组成的多学科团队，开发了远程口腔行为评估系统（ROBAS），该系统可以提供有关口腔卫生行为的客观、个人层面和生态有效的数据。以我们团队开发的强大研究先验和创新技术为基础，我们的学术与产业合作伙伴关系将分三个连续阶段构建、测试、完善和现场验证 ROBAS，并实现相应的具体目标。在第一阶段，我们将通过将支持传感器的牙刷与腕部活动记录仪和数据分析系统集成来开发 ROBAS，然后使用计算机模拟器和测试对象进行基准测试和迭代完善（目标 1）。在第二阶段，我们将在 32 名健康受试者中对 ROBAS 进行试点测试，以优化其在自然环境中的性能和可用性（目标 2）。收集的客观（精度、有效性、可靠性）、主观（用户满意度、可接受性）和用户性能参数的数据将用于进一步完善 ROBAS 的有效性、可靠性和功能。在第三阶段，我们将使用一项随机、对照、交叉研究，将 60 名高危（物质使用）个体与 60 名可比对照进行对比，以评估 ROBAS 在自然环境中纵向（6 个月）评估口腔健康行为 (OHB) 的能力。具体来说，我们将测试 ROBAS 的临床实用性和相对于回顾性自我报告和生态瞬时评估 (EMA) 的增量有效性（目标 3）。此外，我们将使用明确的理论框架来研究社会行为决定因素、OHB（通过自我报告、EMA 和 ROBAS 衡量）和牙科疾病指标（牙龈炎和牙菌斑）之间的关联（目标 4）。在具有不同牙科疾病风险水平的社区受试者（物质使用者和非使用者）中评估 ROBAS 将有助于我们研究的可转化性。紧密协调的学术产业联盟将确保 ROBAS（项目可交付成果）的快速产品化和可用性。长期收集精细的、生态有效的数据的能力将为开发和严格测试更复杂的口腔健康行为因果模型提供前所未有的机会，并收集有关行为改变过程的基本见解。从传统的救援驱动的牙科护理模式到基于传感器的行为分析的主动预防方法的范式转变有可能从根本上改变牙科护理的提供。同样重要的是，我们项目的技术创新将在各种社会、行为和临床环境中得到广泛应用。

 DESCRIPTION (provided by applicant): Dental disease is one of the most common chronic diseases in the United States and exacts a substantial personal and societal toll. A large body of research links dental disease to poor oral hygiene behaviors such as inadequate brushing and flossing, but fails to explain adequately how these putative causal relationships arise or are mediated. Traditional retrospective self-reports are susceptible to numerous biases and lack the granularity and scientific rigor needed to rigorously test hypotheses about causal pathways. To solve the vexing challenges of oral behavior measurement, we have assembled a multidisciplinary team of scientists and engineers to develop a Remote Oral Behaviors Assessment System (ROBAS) that can provide objective, individual-level and ecologically-valid data on oral hygiene behaviors. Building on a strong body of research priors and innovative technologies developed by our team, our academic-industry partnership will build, test, refine and field-validate ROBAS in three sequential stages with their corresponding specific aims. In Stage 1, we will develop ROBAS by integrating a sensor-enabled toothbrush with a wrist actigraph and a data analytics system and then bench-test and iteratively refine it using computerized simulators and test subjects (Aim 1). In Stage 2, we will pilot-test ROBAS in a cohort of 32 healthy subjects to optimize its performance and usability in naturalistic settings (Aim 2). Data collected on the objective (precision, validity, reliability), subjective (user satisfaction, acceptability) and user performance parameters will be used to further refine ROBAS validity, reliability and functionality. In Stage 3, we will use a randomized, controlled, crossover study contrasting 60 at-risk (substance using) individuals with 60 comparable controls to evaluate ROBAS's capacity for longitudinal (6 months) assessment of oral health behaviors (OHBs) in naturalistic settings. Specifically, we will test ROBAS's clinical utility and incrementa validity over retrospective self-reports and Ecological Momentary Assessments (EMAs) (Aim 3). Additionally, we will use an articulated theoretical framework to investigate the associations among sociobehavioral determinants, OHBs (measured by self-reports, EMAs and ROBAS), and dental disease indicators (gingivitis and dental plaque) (Aim 4). Evaluating ROBAS in community subjects with different levels of risk for dental disease (substance users and non-users) will facilitate translatability of our research. The tightly coordinated academic-industry alliance will ensure rapid productization and availability of ROBAS (project deliverable). The ability to collect granular, ecologically-valid data for long periods of time will afford unprecedented opportunity to develop and rigorously test more sophisticated causal models of oral health behaviors and glean fundamental insights regarding behavior change processes. The paradigm shift from the conventional rescue- driven model of dental care to a proactive preventive approach, based on sensor-enabled behavioral analysis, has the potential to fundamentally transform the delivery of dental care. Equally important, the technological innovations of our project will have broad application in a variety of social, behavioral and clinical settings.