Machine Learning Approaches in Label-Free Small Extracellular Vesicles Analysis with Surface-Enhanced Raman Scattering (SERS) for Cancer Diagnostics

Der Vang,Maria Kelly,Manju Sharma,Manisha Sheokand,Leyla Esfandiari,Ruxandra Ioana Dima,Pietro Strobbia
DOI: https://doi.org/10.1101/2024.02.19.581099
2024-06-01
Abstract:Early diagnosis remains of pivotal importance in reducing patient morbidity and mortality in cancer. To this end, liquid biopsy is emerging as a tool to perform broad cancer screenings. Small extracellular vesicles (sEVs), also called exosomes, found in bodily fluids can serve as important cancer biomarkers in these screenings. Our group has recently developed a label-free electrokinetic microchip to purify sEVs from blood. Herein, we demonstrate the feasibility to integrate this approach with surface-enhanced Raman scattering (SERS) analysis. SERS can be used to characterized extracted sEVs through their vibrational fingerprint that changes depending on the origin of sEVs. While these changes are not easily identified in spectra, they can be modeled with machine learning (ML) approaches. Common ML approaches in the field of spectral analysis use dimensionality reduction method that often function as a black box. To avoid this pitfall, we used Shapley additive explanations (SHAP) is a type of explainable AI (XAI) that bridges ML models and human comprehension by calculating the specific contribution of individual features to a model predictions, directly correlating model/decisions with the original data. Using these approaches we demonstrated a proof-of-concept model predictive of cancer from isolated sEVs, integrating the electrokinetic device and SERS. This work explores the use of explainable AI to perform diagnostic analysis on complex SERS data of clinical samples, while reporting interpretable biochemical information.
Bioengineering
What problem does this paper attempt to address?