PFI-AIR-TT: Wearable sleepwear for quantitative prognostication and noninvasive therapy of obstructive sleep apnea

该PFI：AIR技术翻译项目专注于翻译先进的预测和预测方法，以满足对治疗睡眠障碍的经济有效的家庭解决方案的需求。阻塞性睡眠呼吸暂停（OSA）等睡眠障碍在全国人口中估计占6.6%。睡眠治疗市场每年超过230亿美元，并将以每年15%以上的速度增长。 然而，目前的处理程序有严重的缺点，并在消费者论坛上受到高度批评。传感器和无线通信的最新进展为治疗睡眠呼吸暂停的经济高效的家庭护理选择提供了前所未有的机会。该项目将产生一个抓住这一机会的SleepEaze技术原型。这项技术是基于使用无线生物识别睡衣来监测呼吸暂停患者在睡眠期间的生物节律，它采用了一种正在申请专利的算法来预测呼吸暂停事件发生前几分钟。然后，当预测事件即将发生时，可以刺激负责阻塞的肌肉（“扼杀在萌芽状态”），让患者获得更宁静的睡眠。创新的预测算法和低功耗电子元件的使用为无创刺激提供了突破，可以产生一种新型的可负担的可穿戴设备，用于治疗睡眠呼吸暂停，具有很高的商业化潜力。该项目计划追求三个主要目标：（a）技术改进，包括：改进可穿戴传感器布局以增强舒适度，基于非参数狄利克雷过程高斯过程混合表示来调整预测模型的参数以减少与多步前瞻程序相关的计算开销，以及使用新颖的低功率电子器件来优化刺激模态、位置和时间-频率刺激分布，以有效地避免睡眠呼吸暂停事件;（B）使用早期原型的扩展测试，将正在进行的人类受试者研究扩展到基准信号质量、预测和刺激性能，和睡眠质量的改善，从技术准备批准（510 K）从联邦监管机构;以及（c）用于实时治疗睡眠呼吸暂停发作的更新的可穿戴多传感器单元的原型制作，以及其测试和基准测试。如果成功，拟议的项目将产生一个功能完善的SleepEaze设备原型（硬件和相关软件接口），为OSA治疗提供一种全新的方法。此外，参与该项目的人员，包括一名博士后助理和两名研究生，将获得通过得克萨斯A M大学和NSF西南I军团节点提供的创业经验。&

This PFI: AIR Technology Translation project focuses on translating an advanced prediction and prognostics approach to fill the need for affordable and effective at-home solutions to treat sleep disorders. Sleep disorders such as obstructive sleep apnea (OSA) are prevalent in an estimated 6.6% of the national population. The market for sleep treatment is noted to exceed $23 billion per year, and is set to grow at over 15% annually. However the current treatment procedures suffer from severe drawbacks and are highly criticized in consumer forums. Recent advancements in sensors and wireless communication offer an unprecedented opportunity for cost-effective home care options to treat sleep apnea. This project will result in a prototype of SleepEaze technology that captures this opportunity. This technology is based on using a wireless biometric sleepwear to monitor the apnea patient's biorhythms during sleep, and it employs a patent-pending algorithm to predict apnea events several minutes ahead of their onset. Then the muscles responsible for the obstruction can be stimulated when the event is predicted to be imminent ("nip in the bud"), allowing the patient a more restful sleep. The use of the innovative prediction algorithm as well as low power electronic elements offer a breakthrough in the noninvasive stimulations that can lead to a new class of affordable wearable devices for sleep apnea treatment with high commercialization potential.This project plans to pursue three major objectives: (a) Technology Improvement, which includes: the refinement of the wearable sensor layout to enhance comfort, tuning of parameters of the prediction model based on nonparametric Dirichlet Process Mixture of Gaussian Process representations to reduce computational overhead associated with multi-step look ahead procedures, and optimization of the stimulation modality, location, and time-frequency stimulation profiles using a novel low power electronics to effectively avert sleep apnea events; (b) Extended testing using an early prototype, extending an ongoing human subject study to benchmark signal quality, prediction and stimulation performance, and sleep quality improvement from the technology in preparation for approvals (510K) from the federal regulatory agencies; and (c) Prototyping of a updated wearable multi-sensor unit for real-time treatment of sleep apnea episodes, as well as its testing and benchmarking. If successful, the proposed project will result in a functional, refined prototype of the SleepEaze device (hardware and associated software interface) that offers a radically new approach for OSA treatment. In addition, personnel involved in this project, including a post-doctoral associate and two graduate students, will receive entrepreneurship experiences offered through Texas A&M University and the NSF Southwest I-Corps node.