ASCENT: Multimodal chest e-tattoo with customized IC and deep learning algorithm for tracking and predicting progressive pneumonia

Coronavirus infections may cause life-threatening pneumonia with a mortality rate more than 10% in certain populations, which could quickly overwhelm any medical care system. Continuous monitoring of the infected and suspected at the hospital or under self-quarantine can help optimize triage and treatment. However, so far there is no available mobile device and algorithm platform that can perform reliable, comprehensive, continuous and long-term monitoring and assessment for pneumonia patients in either clinical or free-living environments. The goal of this ASCENT research is to develop, integrate, and test foundational technologies required for a scalable monitoring and triage system for patients who have contracted pneumonia. The objective is to integrate a wireless, noninvasive, week-long wearable, and multimodal physiological sensor platform (e-tattoos) with a dedicated integrated circuit (IC), connect it to an FDA (U.S. Food and Drug Administration) cleared virtual patient monitoring platform (Sickbay) which also hosts a customized deep learning algorithm, for the continuous monitoring and assessment of the severity of progressive pneumonia. The result will be a gamechanging hardware and software system that provides continuous monitoring and intelligent assessment for highly-infectious and critically-ill patients but also protects healthcare providers from infection and contamination.

There is a longstanding systems challenge that the world lacks long-term, high-fidelity, continuous and scalable clinical surveillance platforms for infectious disease patients to battle with global pandemic like COVID-19. The progression of pneumonia is associated with the changes in vital signs such as core body temperature, respiratory rates, heart rates, blood oxygen saturation and so on. Since clinical deterioration of patients at risk of developing pneumonia can be short and unpredictable, continuous multimodal monitoring and accurate assessment is necessary for this population, whether in the hospitals or at home. The five investigators bring together well-established expertise in multimodal wearable sensors (Lu), mixed signal IC design (Li), time-series data analytics (Miao), clinical systems integration and scalable patient monitoring (Rusin), as well as critical care medicine (Jain). This multidisciplinary engineering and clinical team attempt to address this system-level challenge through: 1) development of wireless wearable sensors called e-tattoo with dedicated IC capable of noninvasive and week-long multimodal patient monitoring; 2) data analysis and deep learning algorithm development and integration with e-tattoo through an FDA (U.S. Food and Drug Administration) cleared virtual patient monitoring platform, Sickbay; 3) e-tattoo and algorithm validation on 20 patients with progressive pneumonia at Texas Children's Hospital. The broader impacts for the society are dramatically improving how critically ill patients are monitored as well as training next generation engineers to carry out convergent research. The ultimate vision is to establish a scalable means of safely surveilling patients and orchestrating high-quality care across the country.

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.