Date of Award
Master of Science (MS)
Inflammatory skin diseases are estimated to impact 20% of the global population and are the fourth leading cause of nonfatal disability worldwide. Diagnosis and management are predominantly based on clinician visual assessment of disease related changes in skin morphology. The qualitative nature of this method can result in misdiagnosis and underdiagnoses of treatable diseases. There persists systematic undertreatment of skin of color patients given the more subtle presentation of erythema against pigmented skin. There exists a clinical need for a quantitative and objective inflammation assessment tool that meets the needs of a diverse patient population. The accumulation of interstitial fluid is a pathophysiological hallmark of inflammation. Diffuse reflectance imaging in the short-wave infrared (SWIR) spectrum is an effective method for quantitatively mapping relative cutaneous fluid content. Unlike the visible and near infrared spectra, the SWIR spectrum exhibits strong water absorption and relatively weak melanin absorption. Therefore, SWIR affords the opportunity to obtain high contrast images of edema in a manner unaffected by skin tone. Images can be used to derive improved visualizations and quantitative measures of the inflammatory state beyond what is visible to the clinician. The lack of a suitable clinical imaging system has limited further validation of this application. To address this need, a novel mobile multispectral wide-field imaging modality is presented. This thesis primarily elaborates on component selection, integration, and characterization of the imaging system and accompanying image analysis pipeline. A single subject model is used to demonstrate the ability of this technology to capture inflammatory processes beyond what is visible to the human eye. Additionally, we discuss the device's compatibility with clinical data vii acquisition and priorities for next generation designs. This work also provides a preliminary evaluation of the impact of tissue optical properties on the SWIR water signal. Tissue-mimicking phantoms demonstrate that the signal remains strong through several millimeters of scattering tissue and is not heavily influenced by the presence of melanin mimicking optical absorbers. This work represents a critical step toward facilitating subsequent validation studies that demonstrate the clinical utility of multispectral SWIR imaging.
Dr. Leonid Shmuylovich
Dr. Christine O’Brien Dr. Mikhail Berezin
Available for download on Thursday, May 16, 2024