Chemometrics-aided Surface-enhanced Raman spectrometric detection and quantification of GH and TE hormones in blood

TL;DR

This study demonstrates that Surface-Enhanced Raman Spectroscopy combined with artificial neural networks can rapidly and accurately detect and quantify GH and TE hormones in blood, offering a promising tool for biomedical and sports science applications.

Contribution

The paper introduces a novel SERS-ANN method for hormone detection in blood, achieving high accuracy and rapid analysis, with potential for non-invasive diagnostics.

Findings

High accuracy in hormone quantification with R^2 > 87.71%

Rapid analysis time of approximately two minutes

SERS provides a simple, non-specific detection method

Abstract

This work explores the use of Surface-Enhanced Raman Spectroscopy (SERS) combined with artificial neural network (ANN) models to detect and quantify growth hormone (GH) and testosterone (TE) in the blood of Sprague Dawley (SD) rats. SERS spectra were recorded from blood samples of SD rats injected with GH, TE, both hormones, and non-injected controls using 785 nm laser excitation. The samples were mixed with silver nanoparticles (AgNPs) synthesized in distilled water, applied onto a microscope slide, and air-dried. The resulting SERS spectra displayed similar profiles with intensity variations depending on the hormone, revealing specific bands at 658, 798, 878, 914, 932, 1064, 1190, 1354, 1410, and 1658 cm-1. PCA analysis indicated time-dependent intensity changes in bands centered around 1378 (all groups), 658 and 1614 cm-1 (GH-injected rats), and others for different hormone combinations. These variations reflect subtle biochemical changes induced by hormone injections. The ANN models, trained with six PCA scores of blood spiked with various hormone concentrations, showed high accuracy, with coefficients of determination greater than 87.71% and low root mean square error (RMSE) values below 0.6436. The hormone levels in injected rats increased initially and later declined, a trend confirmed by ELISA kits. Although ELISA and SERS produced similar results, SERS offered advantages such as rapid analysis (about two minutes), simple sample preparation, small sample volumes, and non-specificity to hormones. This suggests that SERS, combined with ANN models, could be used to detect exogenous sports dopants. These findings expand the potential applications of SERS in sports science, clinical diagnostics, and biomedical research.