Raman Hyperspectroscopy of Saliva and Machine Learning for Sjögren’s Disease Diagnostics

Sjögren's syndrome is an autoimmune disorder affecting exocrine glands, causing dry eyes and mouth and other morbidities. Polypharmacy or a history of radiation to the head and neck can also lead to dry mouth. Sjogren's syndrome is often underdiagnosed due to its non-specific symptoms, limited awareness among healthcare professionals, and the complexity of diagnostic criteria, limiting the ability to provide therapy early. Current diagnostic methods suffer from limitations including the variation in individuals, the absence of a single diagnostic marker, and the low sensitivity and specificity, high cost, complexity, and invasiveness of current procedures. Here we report on utilizing Raman hyperspectroscopy (RHS) combined with machine learning to develop a novel screening test for Sjögren's syndrome.
RHS in combination with machine learning emerges as a powerful and non-destructive technique for studying micro-heterogeneous systems, particularly body fluids. The method is based on generating the hyperspectral data cube like that in hyperspectral imaging. However, the analysis of the data targets obtaining information about the composition of the sample and not restoring the image. As a result, RHS offers a comprehensive understanding of biochemical components of samples opening up practical applications in disease diagnostics. The unique capability of RHS lies in its capacity to integrate the contribution of multiple biomarkers into a spectroscopic signature. However, the subtle spectral changes caused by different components can be challenging to establish visually. To address this, machine learning becomes instrumental. In this study, the method effectively differentiates Sjögren's syndrome patients from healthy controls and radiation therapy patients. This technique shows potential for developing a single non-invasive, efficient, rapid, and inexpensive medical screening test for Sjögren's syndrome using a Raman hyperspectral signature.