SBIR Phase I: A smart wearable platform for remote respiratory monitoring: building better technologies for telemedicine in the age of COVID-19

这项小企业创新研究（SBIR）第一阶段项目的更广泛影响/商业潜力是创建一个智能可穿戴听诊器平台，作为远程监测受COVID-19影响的患者的新工具。许多受感染的患者可能直到为时已晚才出现症状。在出现呼吸窘迫之前远程监测这些患者咳嗽和呼吸短促的发展是至关重要的。患有现有心肺疾病的患者因COVID-19感染病毒性或继发性细菌性肺炎的风险增加，但由于医护人员接触的风险，持续评估这些患者的肺音具有挑战性。显然，需要更有效的方法来监测因COVID-19而被隔离的患者和急性护理患者的呼吸健康状况。该项目允许远程监测，以帮助对COVID-19患者进行分类，并减少医护人员接触。这项小型企业创新研究（SBIR）第一阶段项目旨在进一步开发和优化一种基于人工智能的可穿戴设备，该设备可在高环境噪声环境中监测和分析肺部声音。环境噪音会影响标准电子听诊器的使用。许多市售电子听诊器通过减小动态范围或警告用户不要在高噪声环境中使用设备来解决环境噪声问题。这些缓解方法通过限制可从声学测量中获得的信息，限制了这些设备的效用。此外，对环境噪声的敏感性消除了它在家庭环境中的潜在用途。环境噪声已被证明会降低在低噪声环境中训练的机器学习算法的有效性，从而无法准确检测肺部声音。该项目使用新颖和成熟的被动噪声消除技术、主动噪声消除技术、信号处理技术和机器学习算法来解决高环境噪声问题。这些解决方案在可穿戴呼吸监测平台中的最佳组合和集成将建立一个在各种现实环境中使用的有用工具。本项目的成功与否将通过系统优化后机器学习灵敏度指标的提高来衡量。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

The broader impact /commercial potential of this Small Business Innovation Research (SBIR) Phase I project is to create a smart wearable stethoscope platform as a new tool to remotely monitor patients affected by COVID-19. Many infected patients may not present with symptoms until it is too late. Remotely monitoring these patients for the development of cough and shortness of breath prior to presentation in respiratory distress is critical. Patients with existing cardiopulmonary disease are at increased risk of contracting viral or secondary bacterial pneumonia due to COVID-19, but it is challenging to continuously assess these patients’ lung sounds due to risks of healthcare worker exposure. There is a clear need for more effective ways to monitor patients’ respiratory health due to COVID-19 both in quarantined patients and those in acute care. This project allows for remote monitoring to help triage COVID-19 patients and reduce healthcare worker exposure.This Small Business Innovation Research (SBIR) Phase I project addresses the further development and optimization of an artificial intelligence-based wearable device that monitors and analyzes lung sounds in high ambient noise environments. Ambient noise affects the use of standard electronic stethoscopes. Many commercially available electronic stethoscopes address ambient noise by reducing dynamic range or by warning the user not to use the device in a high noise environment. These mitigation methods restrict the utility of these devices by limiting the information that can be obtained from the acoustic measurements. Additionally, susceptibility to ambient noise eliminates its potential use in the home environment. Ambient noise has been shown to degrade the effectiveness of machine learning algorithms trained in low-noise environments to accurately detect lung sounds. This project addresses issues with high ambient noise using novel and established techniques of passive noise cancellation, active noise cancellation, signal processing techniques, and machine learning algorithms. The optimal combination and integration of these solutions in a wearable respiratory monitoring platform will establish a useful tool for use in a variety of real-world environments. The success of this project will be measured by the improvement of the machine learning sensitivity metrics after system optimization.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.