Wearable alert system for detecting postoperative hypotension

项目摘要 该项目旨在开发一种低成本，舒适，易于使用的可穿戴警报系统， 非侵入性地跟踪患者的连续BP， 术后设置。术后低血压（POH），当患者的平均动脉压 (MAP)福尔斯下降到不安全的水平，通常<70 mmHg，可能经常发生在 离开手术室和出院前。POH是一种严重而常见的 已被证明与不良患者结局独立相关的疾病， 如急性肾损伤、中风、再次入院、心肌损伤和死亡。1 -9期间 在麻醉后监护室（PACU）或重症监护室（ICU）进行手术和恢复， 密切监测患者的血流动力学，以支持及时的干预和治疗。 然而，一旦患者被转移到较低的病房进行恢复， 急剧减少。目前在普通病房监测病人血压的方法， 依赖于护士每4-6小时手动进行的间歇性抽查， 血压计袖带不足以检测POH事件。最近的一项研究表明， 50%的过敏事件未被常规生命体征评估检测到。存在未满足的 需要一种在术后环境中舒适地监测患者血压的技术。我们 核心技术能够在任何位置非侵入性地测量快速BP变化， 可触知的脉搏拟议项目旨在建立在核心技术的基础上， 警报系统，告知护理人员术后的痉挛发作。为此，一台机器 将通过以下方式开发用于检测癫痫发作的ML分类器： ICU中的监控。该ML分类器将在PACU中进行可行性测试，然后进行 在普通病房进行试点研究。成功完成拟议目标将证明- 概念可穿戴警报系统，可在术后环境中持续监测BP 以一种低调、无线和舒适的方式。这种技术具有巨大的潜力， 通过更早地检测患者病情恶化来改变临床实践， 干预，最终改善患者的预后。

Project Summary This project aims to develop a low-cost, comfortable, and easy-to-use wearable alert system that tracks continuous BP non-invasively in patients to notify caregivers of hypotensive events in the postoperative setting. Postoperative hypotension (POH), when a patient’s mean arterial pressure (MAP) falls to unsafe levels, commonly <70 mmHg, can occur frequently in the time span between leaving the operating room and prior to discharge from the hospital. POH is a serious and common condition that has been shown to be independently associated with poor patient outcomes such as acute kidney injury, stroke, hospital re-admission, myocardial injury, and death.1–9 During surgery and recovery in the post anesthesia care unit (PACU) or the intensive care unit (ICU), a patient’s hemodynamics are closely monitored to support timely interventions and treatment. However, once the patient is transferred to a lower ward for recovery, patient monitoring is dramatically reduced. The current method to monitor a patient’s BP in the general ward, which relies on intermittent spot-checks performed manually every 4-6 hours by the nurse using an oscillometric BP cuff, is insufficient for detecting POH events. A recent study showed that almost 50% of hypotensive events went undetected by routine vital assessments. There is an unmet need for a technology that comfortably monitors a patient’s BP in the postoperative setting. Our core technology is able to non-invasively measure rapid BP changes at any location with a palpable pulse. The proposed project aims build upon the core technology to create a wearable alert system that informs caregivers of postoperative hypotensive episodes. To do so, a machine learning (ML) classifier for detection of hypotensive episodes will be developed through monitoring in the ICU. This ML classifier will undergo feasibility testing in the PACU and then a pilot study in the general ward. Successful completion of the proposed aims will result in a proof- of-concept wearable alert system that continuously monitors BP in the postoperative environment in a low-profile, wireless, and comfortable manner. Such a technology has huge potential to change clinical practice through earlier detection of patient deterioration, allowing more timely intervention and ultimately improved patient outcomes.