RAPID: Screening and Prognosis of COVID-19 by a Novel RF Stethoscope

虽然新冠疫情给我们的社会带来了灾难，但作为第一道防线的公共筛查技术仍然薄弱。仅仅发烧是不够的；呼吸困难、胸痛、嗅觉丧失依赖于自我报告，因此不可靠或不准确。但是，如果没有有效和快速的呼吸道传染病筛查，就很难恢复正常的社会交往，而不严重担心新一波COVID-19感染或新冠病毒造成同等或更大的损害。COVID-19症状观察的来源是肺炎。组织炎症引起发烧和嗅觉丧失；肺组织积水引起的充血会导致呼吸困难和胸痛。人体呼吸，像许多其他身体功能一样，会在组织受损的情况下自我补偿，这使得在目前的筛查标准下，早期的肺炎是“无症状的”。本项目旨在开发和测试一种有效的传感器技术，利用射频（RF）信号检测肺组织损伤。这项技术借鉴了用百年历史的听诊器听呼吸声音的想法。肺部和气道组织受损会引起异常声音或不同的组织振动特征，这可以迅速识别为COVID-19感染的另一个迹象。然而，传统的听诊器需要裸露的皮肤接触，以获得高质量的记录和安静的环境，以尽量减少干扰。该项目将开发一种新的射频听诊器，它可以测量衣服上的组织振动特性，并且可以在公共交通等非常嘈杂的环境中可靠地工作。该研究将建立一个健康和确诊COVID-19患者的呼吸特征数据库。基于成熟的无线技术，无线听诊器可以廉价而广泛地部署。用户可以到检查点，在几秒钟内完成筛查，以便进行COVID-19的早期筛查。然后可以进一步进行病理检查和社会隔离，完成对这次大流行的必要管理。这种新的射频听诊器技术提供了一种快速方便的方法来测量肺组织振动特性，并可应用于与心肺功能相关的其他医疗保健应用。本项目旨在通过新型射频听诊器建立COVID-19症状筛查和持续预后平台，该听诊器可作为公共交通工具的快速和具有成本效益的筛查，也可作为医疗点设施或家庭的复发预后工具。所提出的方法中的传感模式是肺组织振动特性，它将在肺炎的所有阶段发生变化。该方法与传统的声学听诊器相似，但读出的是射频信号而不是声音信号。通过衣物上的射频传感探头，将少量电磁能量耦合到气管、支气管和肺实质中，测量组织振动特征，通过组织粘弹性和含水量来区分健康人和潜在的COVID-19患者。主要研究任务包括射频传感器的开发、健康成人的测试、对新冠肺炎潜在患者的长期持续监测等。通过听诊生理学和机器学习方法将COVID-19感染特征与正常呼吸数据库区分开来。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 COVID-19 pandemic has brought disasters to our society, but the public screening technology as our first line of defense remains weak. Fever alone is not sufficient; difficulty in breathing, chest pain and anosmia rely on self report, and hence not reliable or accurate. However, without effective and fast screening against respiratory contagious diseases, it is difficult to resume normal social contacts without serious concerns of a new wave of COVID-19 infection or a new coronavirus that causes equal or more damages. The source of symptomatic observation of COVID-19 is pneumonia. The tissue inflammation causes fever and anosmia; the congestion from water in lung tissues causes difficulty in breathing and chest pain. Human respiration, like many other body functions, will self compensate in the presence of damaged tissues, which makes the early stage of pneumonia “asymptomatic” under the present screening criteria. This project seeks to develop and test an effective sensor technology to detect the lung tissue damage using radio-frequency (RF) signals. The technology borrows the ideas of listening to breathing sounds by century-old stethoscopes. Damaged tissues in the lung and airway will cause abnormal sounds, or different tissue vibration characteristics, that can be identified quickly as another indication of COVID-19 infection. However, conventional stethoscopes require bare skin contact for quality recording and quiet ambience to minimize interference. The project will develop a new RF stethoscope, which can measure tissue vibration characteristics over clothing and can operate reliably in very noisy environment such as public transits. The research will build a database of breathing features from healthy and confirmed COVID-19 patients. Based on the mature wireless technologies, the radio stethoscope can be inexpensively and broadly deployed. The user can go to a checkpoint and finish the screening within a few seconds for early screening of COVID-19. Further pathological tests and social isolation can then follow through the necessary management of this pandemic. This new RF stethoscope technology provides a fast and convenient way to measure lung tissue vibration characteristics and can be applied to other healthcare applications related to cardiopulmonary functions.This project aims to establish a symptomatic screening and continuous prognosis platform for COVID-19 by a novel RF stethoscope, which can be deployed as a fast and cost-effective screening in public transit and as a recurrent prognosis tool in point-of-care facilities or at home. The sensing modality in the proposed method is the lung tissue vibration characteristics which will change in all stages of a pneumonia. The method is similar to that of the conventional acoustic stethoscope, but the readout is by the RF signals instead of the sound signals. A small amount of the electromagnetic energy will be coupled deeply into the trachea, bronchus and lung parenchyma by the RF sensing probes over clothing, and the tissue vibration features will be measured to distinguish healthy persons from potential COVID-19 patients by the tissue viscoelasticity and water content. The major research tasks include the RF sensor development, testing on healthy adults, and long-term continuous monitoring of potential COVID-19 patients at the Weill Cornell Medical Center in New York City. Features for COVID-19 infection will be distinguished against the normal breathing database by auscultation physiology and machine learning methods.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.