Clinical Diagnosis of Diabetes Using Surface-enhanced Raman Spectroscopy Liquid Biopsy and Machine Learning

Overview

This project aims to adapt the gold nanoparticle-based surface-enhanced Raman spectroscopy (SERS) technology to clinical application. In this exploratory study, a measurement protocol will be established to investigate whether SERS (combined with multivariate data analysis or machine learning algorithms) allows the diagnosis of patients with diabetes.

This study on the clinical application of surface-enhanced Raman spectroscopy (SERS) comprises two parts. First, a SERS measurement protocol will be developed to enhance the interactions between gold nanoparticles and the components of the patient's samples, maximizing Raman spectroscopical signatures. Given the complex composition of human blood, which encompasses numerous biological constituents, the study focuses on serum, a component obtained through centrifugation after removing cells and clotting factors. Fifteen spectra will be recorded per sample. The raw spectra will be post-processed, including removal of the substrate signal, baseline correction, vector normalization, and smoothing steps. The SERS measurement protocol established in the first section will subsequently be applied to samples of healthy and diabetes patients. Two different approaches will be followed. First, multivariate data analysis will be performed to identify distinctive feature characteristics in the samples that correlate to their group (healthy and diabetes patients), allowing patient diagnosis. Second, different machine learning algorithms and data augmentation strategies will be explored for better patient diagnosis.

Conditions

Interventions

Eligibility

Locations (1)

Outcomes