Wearable Multi-modality Cuffless Blood Pressure Monitoring

摘要 连续血压(BP)是麻醉、手术中最重要的监测参数之一 以及重症监护病房(ICU)。低血压和高血压都会损害重要器官的功能。 (例如，脑、心脏和肾脏)，术中低血压与术后死亡率有关， 因此，尽快检测到血压变化，以及时进行干预或治疗非常重要。 然而，目前血压监测的黄金标准技术--有创动脉(A-line)--会导致 患者遭受(身体疼痛)，并增加感染的风险。在美国，大约8万人的血液 由动脉导管引起的溪流感染每年都有报告。由于与此相关的固有风险 对于a-line，它只适用于临床显示的高风险手术或ICU患者。作为A线的结果 风险和不适，尽管全球每年进行超过3亿次手术，但只有 一小部分人接受持续的血压监测。此外，虽然生命体征(心电、脉搏血氧饱和度、血压等) 监护是手术室和ICU的常规工作，目前大多数监护设备都是个人安装的 房间，这会导致患者监控的空白、患者运送过程中的事故以及额外的工作 离开和进入新设施时，断开并重新连接传感器。无缝的“连续护理” 监控--例如从手术室到ICU，包括在两者之间和不重新连接的情况下进行运输 传感器-是临床医生最想要的东西。近年来，人们一直在努力开发便携式 心电监护仪和“移动ICU”；然而，到目前为止，还没有连续和无缝的血压监测 已实现。该提案充分利用了相关R21(EB022271)项目的成果。我们会 开发基于机器学习和深度学习的新型数据融合算法，以利用现有的生命体征 持续的血压监测，然后将它们与我们独特的可穿戴式患者监测系统集成在一起，形成 新型围手术期病人监护系统。我们将对照黄金标准a测试系统的性能- Line和Finapres BP技术公司。开发一种功能齐全的技术，用于无创、连续和 无缝的BP监测。我们还将为未来的BP技术开发开发一个公共数据库。这个 拟议的多模式算法、无缝BP监测系统和PhysioNet数据库将提供 为满足非侵入性连续血压监测的临床需求而采取的主要步骤。

Abstract Continuous blood pressure (BP) is one of the most critical monitoring parameters during anesthesia, surgery and in intensive care units (ICU). Both hypotension and hypertension can impair the function of vital organs (e.g. brain, heart and kidneys), and intraoperative hypotension is associated with postoperative mortality, which makes it important to detect BP changes as quickly as possible to prompt timely intervention or therapy. However, the current gold standard technology for BP monitoring, an invasive arterial line (a-line), causes patient suffering (physical pain) and increases the risk of infection. In the United States, about 80,000 blood stream infections caused by an arterial catheter are reported annually. Due to the inherent risks associated with a-line, it is used only for clinically indicated high risk surgeries or ICU patients. As a result of the a-line risks and discomforts, even though more than 300 million surgeries are performed worldwide each year, only a small portion receive continuous BP monitoring. In addition, although vital sign (ECG, pulse oximetry, BP etc) monitoring is routine in surgical rooms and ICUs, currently most monitoring devices are fixed in individual rooms, which result in gaps in patient monitoring, accidents during patient transport process, and extra work to disconnect and reconnect sensors when leaving and entering a new facility. Seamless “continuum of care” monitoring—for instance from surgical room to ICUs, including transport in between and without reconnecting sensors—is on top of the wish list by clinician. In recent years, efforts have been made to develop portable ECG monitors and “mobile ICUs”; however so far, no continuous and seamless BP monitoring has been achieved. This proposal fully leverages the outcomes from the related R21 (EB022271) project. We will develop novel machine learning and deep learning based data fusion algorithms to use existing vital signs for continuous BP monitoring, then integrate them with our unique wearable patient monitoring system to form a novel perioperative patient monitoring system. We will test the system’s performance against gold standard a- line and Finapres BP technologies. to develop a fully functional technology for noninvasive, continuous, and seamless BP monitoring. We will also develop a public database for future BP technology development. The proposed multimodality algorithms, seamless BP monitoring system and PhysioNet database will provide major steps forward to meet the clinical need for noninvasive continuous BP monitoring.