Biochemical components of corneal stroma: a study on myopia classification based on Raman spectroscopy and deep...

Area of Science:

• Biochemistry
• Ophthalmology
• Spectroscopy

Background:

• Myopia, a refractive error, involves changes in the eye's structure.
• Understanding biochemical alterations in the cornea is crucial for myopia research.

Purpose of the Study:

• To investigate biochemical differences in corneal stroma lenses among low, moderate, and high myopia.
• To assess the utility of Raman spectroscopy in differentiating myopia severity.

Main Methods:

• Collected Raman spectra from corneal stroma samples of 38 patients undergoing SMILE surgery.
• Categorized patients into low, moderate, and high myopia groups.
• Processed spectral data and employed statistical analyses (t-tests) and machine learning models (PLS-KNN, PCA-KNN).

Main Results:

• Significant differences in Raman spectral peaks related to collagen, lipids, and nucleic acids were observed between myopia groups (P < 0.001).
• The PLS-KNN model achieved high accuracy (95%) in classifying myopia severity.
• Distinguishing between low and high myopia showed excellent performance with PLS-KNN (94.4% accuracy, 0.98 AUC).

Conclusions:

• Raman spectroscopy effectively identifies biochemical variations in the corneal stroma correlating with myopia.
• The findings suggest Raman spectroscopy is a promising tool for myopia pathogenesis research.
• Biochemical signatures in the cornea may serve as biomarkers for myopia progression.