A Multimodal Spectroscopic Approach Combining Mid-infrared and Near-infrared for Discriminating Gram-positive and...

Area of Science:

• Analytical Chemistry
• Spectroscopy
• Microbiology

Background:

• Accurate bacterial identification is vital for public health, addressing antibiotic resistance and food safety.
• Vibrational spectroscopy offers rapid, non-destructive analysis of biological samples.
• Distinguishing Gram-positive and Gram-negative bacteria is a key challenge in clinical and food microbiology.

Purpose of the Study:

• To compare the efficacy of ATR-FTIR and NIR spectroscopy for differentiating Gram-positive and Gram-negative bacteria.
• To investigate the potential of NIR spectroscopy as a low-cost alternative for bacterial identification.
• To explore the benefits of combining ATR-FTIR and NIR data for enhanced bacterial classification.

Main Methods:

• Spectra acquisition from five Gram-positive and seven Gram-negative bacterial species using ATR-FTIR and NIR.
• Application of data fusion and multivariate analysis (PLS-DA, SVM) for spectral classification.
• Validation of classification models using repeated random sampling (100 iterations).

Main Results:

• Both ATR-FTIR and combined ATR-FTIR/NIR models achieved >98% accuracy in differentiating bacterial groups.
• NIR spectroscopy alone demonstrated classification accuracy >90%, highlighting its potential.
• Spectral differences were correlated with distinct bacterial cell wall compositions.

Conclusions:

• ATR-FTIR and NIR spectroscopy are effective tools for rapid bacterial differentiation.
• Low-cost NIR spectroscopy shows significant promise for point-of-care diagnostics in microbiology and food safety.
• Multimodal analysis using both techniques enhances identification reliability and broadens applicability.