Improved detection of head and neck cancer using spatial raman spectroscopy

There are many limitations to the current techniques used for diagnosing head and neck cancer at an early enough stage for successful intervention with an optimal prognosis. Raman spectroscopy (RS) is an imaging technique that can detect low levels of biomarkers at earlier cancer stages than other techniques that require a higher limit of detection. RS is able to inspect the biological composition of a sample and compare the subtle changes in metabolite concentration, creating a biological fingerprint, which provides specific information about the biomarkers present in a sample. Additionally, it is fast and label-free, so it requires no sample preparation. RS was used to measure the molecular composition of plasma that follows the effects of capillary and Marangoni flow, known as the coffee ring effect, as it dries in small droplets. This effect causes smaller particles to dry near the edges, creating a radially symmetric, heterogeneous distribution of molecules across the dried sample that will lead to varied RS spectra according to the location the measurement was collected. This study investigated how we might be able to take advantage of this process in order to gain more insight into the differences between cancerous and healthy samples using RS. We also utilized automation techniques to streamline data collection and lead to reduced need for an expert user and less variability in measurements. By combining these two discoveries, we aim to produce robust and repeatable results that are able to more accurately detect head and neck cancer. Through this research we have found that there is a difference in the intensity of certain molecular bonds across the radius of dried sample droplets, and this leads to varied accuracy, suggesting that the measurements should not be taken at randomly selected locations, but rather at specific distances from the edges.