Quantifying glucose and lipid components in human serum by Raman spectroscopy and multivariate statistics

Area of Science:

• Biomedical Engineering
• Analytical Chemistry
• Spectroscopy

Background:

• Quantitative analysis of human serum biochemical components is crucial for clinical diagnosis.
• Raman spectroscopy offers a non-invasive method for biochemical analysis.
• Developing accurate spectral models is key to leveraging Raman spectroscopy for diagnostics.

Purpose of the Study:

• To develop a spectral model for estimating glucose and lipid fractions in human serum.
• To evaluate the feasibility of Raman spectroscopy for diagnostic purposes.
• To correlate Raman spectral data with established colorimetric assay results.

Main Methods:

• Collected 44 human blood serum samples.
• Performed Raman spectroscopy (830 nm, 250 mW, 50-s accumulation).
• Developed a quantitative model using partial least squares (PLS) regression with leave-one-out cross-validation.

Main Results:

• The PLS model accurately estimated triglyceride (r=0.98) and cholesterol (r=0.96) concentrations.
• Root mean square errors for triglycerides and cholesterol were 35.4 and 15.9 mg/dL, respectively.
• Correlation coefficients (r) for other biochemicals ranged from 0.75 to 0.86.

Conclusions:

• Raman spectroscopy combined with PLS regression can effectively perform biochemical assays in human serum.
• This technique demonstrates potential for routine clinical analysis and diagnosis.
• The study validates Raman spectroscopy as a viable tool for quantitative biochemical analysis.