A Smart Ring for Cuffless Blood Pressure to Reduce Health Disparities in People of Color

项目概要/摘要 高血压是最普遍的心血管危险因素， 在非裔美国人中的结果。患有高血压的非洲裔美国人也更有可能不知道 血压升高，进一步加剧心血管疾病的并发症。因此，检测 准确的血压在非洲裔美国人是一个关键的一步，以减轻高血压不同 结果。该项目的目标是创建一个不显眼的无袖带血压监测器， 测量和识别有色人种社区的隐性高血压。当前状态 结合袖带的技术是麻烦的，并且不允许被动/频繁的测量。的 无袖带解决方案的趋势主要集中在光电体积描记术（PPG）上， 在较深肤色和BMI较高的参与者中也是如此。该项目利用生物阻抗作为 感测模态以指戴式设备的形式，如智能戒指。我们的设备提供了强大的，几乎 连续生物阻抗读数，对应于血容量和血流动力学变化， 通过机器学习和人工智能算法转化为血压。生物阻抗不具有 PPG的局限性;它对不同的肤色不敏感，具有深度渗透能力，因此 较高的BMI（以及皮下脂肪层较厚）不会影响测量结果。的主要重点 我们的研究是为了深入了解隐性高血压，目的是捕获血压 经常在白天和晚上。我们提出的技术可以帮助识别某些短期的 包括血压变化在内的动力学，并允许有效监测对药物的反应 为佐治亚州亚特兰大的非裔美国人和德克萨斯州学院站的非裔美国人和拉丁裔社区提供服务。这 该项目有两个目标。目标1旨在验证已开发的智能戒指技术 在德克萨斯农工大学，在实验室环境中对BMI不同的非裔美国人和拉丁裔参与者（N=60）进行了研究，没有 除了肥胖之外，在存在各种血压调节（冷加压， 三是坚持以人为本，坚持以人为本。目标2扩展了 我们的技术在有和没有高血压史的非裔美国人患者（N=144）中的应用， 亚特兰大，佐治亚州，在混合实验室和门诊环境中，独特的重点是探索报告的现象 和提高有色人种社区的技术接受度和社会相关性。 经过近十年的工业和学术技术发展， 使用PPG，据我们所知，这一努力首次解决了PPG对深色皮肤的有限有效性， 肤色和那些有较高的BMI。文献表明， 非裔美国人社区和肥胖人群中的高血压，这进一步要求BMI 独立的技术开发和验证。

PROJECT SUMMARY/ABSTRACT Hypertension is the most prevalent cardiovascular risk factor and disproportionately responsible for adverse outcomes in African Americans. African Americans with hypertension are also more likely to be unaware of elevated blood pressures, which further exacerbates complications of cardiovascular disease. Thus, detection of accurate blood pressure in African Americans is a critical step towards alleviating hypertension disparate outcomes. The objective of this project is to create an unobtrusive cuffless blood pressure monitor for measurement and identification for masked hypertension in communities of color. Current state of the technology that incorporates cuffs is cumbersome and does not allow for passive/frequent measurements. The trends for cuffless solutions are primarily focused on photoplethysmography (PPG) which does not operate well on darker skin tones and on participants with higher BMIs. This project leverages bioimpedance as a sensing modality in the form of a finger-worn device, like smart rings. Our device provides robust and nearly continuous bioimpedance readings, corresponding to blood volume and hemodynamic changes, that will be translated to blood pressure using machine learning and AI algorithms. Bioimpedance does not have the limitations of PPG; it is insensitive to varying skin tones and has deep penetration capabilities and therefore a higher BMI (and a thicker layer of fat under the skin) does not impact the measurements. The principal focus of our investigation is to gain insights into masked hypertension with the objective of capturing blood pressure frequently during the day and at nighttime. Our proposed technology can help identify certain short-term dynamics inclusive of variations of blood pressure and allows effective monitoring of response to medication for African American in Atlanta, GA and African American and Latinx communities in College Station, TX. This project incorporates two aims. Aim 1 intends to validate the smart ring technology that has been developed at Texas A&M on African American and Latinx participants (N=60) with varying BMI in lab settings, with no known pre-existing condition beyond obesity, in presence of various blood pressure maneuvers (cold pressor, handgrip and exercise) and for three postures (supine, standing and sitting). Aim 2 expands the validation of our technology in African American patients (N=144) with and without a history of hypertension in Atlanta, GA in mixed lab and ambulatory settings with a unique focus to explore the reported phenomenon of masked hypertension and to advance technology acceptance and social relevance in communities of color. After nearly ten years of industrial and academic technology development to extract blood pressure indices using PPG, to our knowledge, this effort for the first time addresses the limited effectiveness of PPG on darker skin tones and those who have a higher BMI. The literature demonstrates the higher prevalence of hypertension in African American communities and among those who have obesity, which further calls for BMI independent technology development and validation.