Introduction:
Gastric cancer remains a major cause of cancer-related mortality worldwide, and early detection is critical for improving patient outcomes. However, current screening methods such as endoscopy are invasive, operator-dependent, and not always suitable for rapid or repeated testing, highlighting the need for objective and minimally invasive diagnostic approaches. This study aimed to develop an acid-resistant surface-enhanced Raman scattering (SERS) platform for direct metabolomic analysis of gastric juice for gastric cancer diagnosis.
Methods:
Polyvinylpyrrolidone (PVP)-encapsulated gold nanoparticles were used to form chemically stable nanocluster films under strongly acidic gastric juice conditions. Gastric juice samples were mixed with the PVP–Au nanoparticle solution and deposited by a meniscus-guided coating method to generate uniform SERS-active films with high reproducibility. The acquired spectra were preprocessed, converted into barcode-format features, and analyzed using a neural-network-based classification model.
Results and Discussion:
A total of 121 gastric juice samples, including 62 from patients with gastric cancer and 59 from controls, were analyzed. The proposed platform achieved high diagnostic performance, with a test-set sensitivity of 91.7%, specificity of 91.7%, and area under the curve of 0.9167. These findings indicate that complex metabolites in gastric juice can be directly profiled without additional pretreatment, while the acid-resistant PVP-coated gold nanocluster structure supports stable SERS measurement and reliable machine-learning-based discrimination. Diagnostic validity was assessed using upper endoscopy with biopsy as the clinical gold standard.
Conclusions:
This study demonstrates that an acid-resistant PVP-coated gold nanocluster SERS platform, combined with neural network analysis, enables accurate gastric cancer detection using gastric juice. The proposed approach may serve as a promising auxiliary diagnostic tool for high-risk screening and clinical prognosis assessment.