Classification of single extracellular vesicles in a double nanohole optical tweezer for cancer detection

Abstract

A major challenge in cancer prognostics is finding biomarkers that can accurately identify cancer at early stages. Extracellular vesicles are promising biomarkers because they: contain cell specific information, are abundant in fluids, and have distinguishing features between cancerous and non-cancerous types. Fluorescent labelling is commonly used to detect extracellular vesicles but has challenges including achieving the desired specificity. Here, we demonstrate a label-free approach to classification of 3 different extracellular vesicle types, derived from non-malignant, non-invasive cancerous, and invasive cancerous cell lines. Using double nanohole optical tweezers, the scattering from single trapped extracellular vesicles is measured, and using a 1D convolutional neural network, we are able to classify the time series optical signal into its respective extracellular vesicle class. This is a promising first step towards early-stage label-free detection of cancers.