A machine learning-driven SERS platform for precise detection and analysis of vascular calcification

Area of Science:

• Biomedical Engineering
• Analytical Chemistry
• Cardiovascular Research

Background:

• Vascular calcification (VC) is a major contributor to cardiovascular disease mortality.
• Current diagnostic and treatment methods for VC are limited.
• Early detection of VC is crucial for effective intervention.

Purpose of the Study:

• To develop a machine learning-assisted surface-enhanced Raman scattering (SERS) technique for label-free analysis of VC.
• To utilize gold nanobipyramid (GNBP) substrates for highly sensitive detection of VC biomarkers in rat serum.
• To establish a new foundation for real-time diagnosis and identification of VC.

Main Methods:

• Preparation of gold nanobipyramid (GNBP) substrates via seed-mediated and liquid-liquid interface self-assembly.
• Measurement of SERS spectra from rat serum samples.
• Application of Principal Component Analysis (PCA)-Linear Discriminant Analysis (LDA) for spectral data processing and classification.

Main Results:

• The developed SERS-assisted machine learning model achieved high performance in classifying VC serum.
• Classification accuracy, sensitivity, and specificity reached 96.0%, with an AUC value of 0.98.
• Key spectral features distinguishing VC were identified through PCA loading plots.

Conclusions:

• The combination of SERS technology and machine learning offers a promising new avenue for the early diagnosis of VC.
• GNBP substrates demonstrate significant advantages for high-sensitivity and high-specificity detection of VC.
• This approach has the potential to significantly improve the early diagnosis and treatment of vascular calcification.