A Highly Sensitive SERS Technique Based on Au NPs Monolayer Film Combined with Multivariate Statistical Algorithms for Auxiliary Screening of Postmenopausal Osteoporosis.

Abstract

Postmenopausal osteoporosis (PMOP) has become an important public health issue. The diagnosis of PMOP relies on clinical symptoms and radiology. However, most patients with PMOP do not exhibit obvious symptoms in the early stages of this disease. This study aimed to explore the feasibility of surface-enhanced Raman scattering (SERS) technology in the auxiliary screening of PMOP. PMOP rats were induced by ovariectomy (OVX) surgery, with a Sham group and an icariin (ICA) treatment group serving as controls. A monolayer film of Au nanoparticles (NPs) was prepared using the Marangoni effect in an oil/water/oil three-phase system, and was used to detect serum SERS signals in the Sham, OVX, and ICA treatment groups. Then, the spectral diagnostic model for PMOP screening was established utilizing partial least squares (PLS) and support vector machine (SVM) algorithms. Histopathology confirmed the establishment of the PMOP rat model. The assignment of Raman peaks and the analysis of spectral differences revealed the biochemical changes associated with PMOP, including the upregulation of tyrosine levels and the downregulation of arginine, tryptophan, lipids, and collagen. When employing the PLS-SVM algorithm to simultaneously classify and discriminate three groups of samples, the diagnostic sensitivity for PMOP is 93.33%, the specificity is 96.67%, and the accuracy of three-class classification is 91.11%. This study demonstrated the potential of SERS for the auxiliary screening of PMOP.