SCH: Smart Breath- based Diagnosis of Pulmonary Exacerbations in Children with Cystic Fibrosis through Machine Learning: Towards Noninvasive Health Monitoring in Real-Time

Cystic fibrosis (CF) is a genetic disorder that negatively affects young people across the globe. Pulmonary exacerbations (PEx), episodes of decreased lung function accompanied by coughing, increased sputum production, and weight loss, are a hallmark of CF lung disease. Besides negative health impacts, PEx

comes with a burden to healthcare costs in the United States and beyond ($10K-$40K USD per episode). Although PEx can be treated with antibiotics, patients still experience decreased quality of life and ultimately reduced survival. A point-of-care, rapid, and accurate test to identify impending PEx that would

benefit from treatment could have an impact on families of children with CF and their clinicians through reducing misdiagnosis and overtreatment. Breath testing, through identifying volatile organic compound (VOC) biomarkers, holds great promise for the development of home-based/clinical testing solutions for

PEx. The goal of this research is to develop a hand-held smart sensor system that can detect exhaled VOC biomarkers for PEx noninvasively in real-time. To accomplish this, machine learning will be utilized to identify a breath-based biosignature of PEx (Aim 1). In parallel, the team will design/test a nanosensor

array to detect the biosignature (Aim 2) and develop a user-friendly smartphone app to be used at-home or in the clinic (Aim 3). Ultimately, this research will further the development of diagnostic solutions for PEx, and advance knowledge in a multi-faceted fashion across disciplines including basic science,

chemistry, engineering, medicine, biotechnology, and health informatics. The technological solution is highly disruptive and challenges the current paradigm of how PEx is diagnosed. From an engineering perspective, sensors are at the cusp of being translated into biomedical devices, and this research aims to

overcome challenges in selectivity that can also be leveraged for VOC-based diagnosis of other heart and lung diseases beyond CF. The interdisciplinary team has vast experience in their respective fields, and their collaboration ensures successful completion of the research. A diverse set of resources/equipment

from the team's laboratories, along with others on campuses, will be leveraged to support research activities. RELEVANCE (See instructions): The research addresses