Project Summary Eosinophilic esophagitis (EoE) is a chronic inflammatory disease affecting the esophagus, and one of the most common causes of vomiting, feeding and swallowing difficulties, including esophageal food impaction in children. EoE is difficult to diagnose and distinguish from other common esophageal diseases in real-time using white light endoscopic imaging of the esophagus conducted via an esophagogastroduodenoscopy (EGD) procedure. Therefore, clinicians rely on multiple random esophageal biopsies which is obtained during this procedure for histologic confirmation (gold standard). However, since EoE is a patchy disease even the multiple biopsies may not yield accurate results. This can lead to a delay in diagnosis and ineffective treatment, and result in preventable complications such as esophageal remodeling requiring surgical interventions. As such there is an urgent need to develop alternative diagnostic tools which can quickly survey more regions of the esophagus, in situ, and provide tissue-specific inflammatory biomarker information for real-time identification of EoE. Raman spectroscopy (RS) is one such technology that can provide a solution as it relies on light interaction with molecules to provide a unique “fingerprint” of the biochemical and molecular composition of a specimen within seconds. Therefore, RS via an endoscope is uniquely suited for real-time biochemical assessment of active EoE. This proposal, presents the development and assessment of a depth specific dual-wavelength Raman endoscope to characterize both biochemical and water changes within the esophagus as it relates to active EoE. To accomplish these goals, Raman biomarkers will be characterized and identified spatially and correlated with known biomarkers of EoE in vivo as well as ex vivo. Computational modeling will be performed using a modified Monte Carlo simulation to optimize probe design based on tissue optical properties. Ex vivo non-linear imaging coupled with Raman maps will be performed on a subset of biopsies, obtained during the EGD procedures, to further study the spatial distribution of key biomarkers and help guide the endoscopic Raman probe design. This probe will be evaluated using optical tissue phantoms and in vivo patient measurements. Finally, a machine learning classification algorithm will be optimized and evaluated to provide a mechanism by which in vivo RS spectra can be classified into active EoE, inactive EoE, GERD (i.e., acid reflux disease) and other non-EoE control. Furthermore, we aim to track patient’s response to treatment and assess Raman spectral changes over time to determine the feasibility of using RS to provide a predictive treatment outcome model. The successful completion of this proposal will yield a novel diagnostic tool for real-time detection and monitoring of EoE in pediatric cases.