This project will devise an integrated sensing and analytical framework to enable non-invasive remote monitoring of in-home/in-hospital patients at risk for sepsis with a wearable diagnostic platform. Sepsis is the leading cause of admitted patient mortality and is one of the most expensive conditions (>$20 billion per year) treated in U.S. hospitals. Early detection and timely intervention are the most crucial factors in improving the outcome for these patients, but the symptoms of sepsis are nonspecific and the repeated blood tests that are required to appropriately monitor are time-consuming and performed at suboptimal rates. These limitations significantly degrade the efficacy of early goal-directed therapy. The core hypothesis of this project is that through the real-time analysis of sweat, sepsis can be detected and its severity can be classified. In that regard, an integrated sensing/analytical framework will be developed to correlate the sweat readings to the sepsis status. Specifically, the research plan includes: 1) devising a chemically-enhanced sweat extraction interface, 2) developing disease-specific inference models and sweat analysis framework, and 3) evaluating the framework applied to septic patients. These efforts will converge toward the development of a patient monitoring platform, which with minimal reconfiguration, can be optimized toward other clinical and physiological applications.
The planned research will address fundamental questions critical to advancing the field of sweat-based biosensing. An integrated sensing technology and analytical framework that normalizes longitudinal sweat readings with respect to multi-conditionally induced secretion parameters will provide quantitative insight into sweat secretion processes and promote our understanding of this complex and multivariate system. Achieving this requires resolving various scientific bottlenecks at multiple fronts spanning from inconspicuous sweat induction at modulated rates, longitudinal/multiconditional in-situ sensing and development and validation of statistical models that can account for the various sources of uncertainty. The team of investigators will capitalize upon their expertise in flexible device fabrication, bio-interface/sensor design, pharmacology, sweat gland physiology, clinical research, and statistical modeling to deliver the proposed objectives. Through in-situ analysis of sweat lactate as a proxy for blood lactate, with a wearable sweat extraction/analysis device, the project can detect the dramatic elevation in blood and detect/classify the severity of sepsis. The outreach component of this project will focus on promoting the STEM participation of underrepresented minorities at high school and undergraduate levels by creating a diverse, STEM-motivated community. Moreover, the educational component of this project will integrate findings from the planned research activities into the course materials.
