Handheld near-infrared spectroscopy for detecting staphylococcal inoculation-associated changes in ex vivo human bone

Area of Science:

• Biomedical Optics
• Spectroscopy
• Orthopedic Research

Background:

• Diagnosing bone infections is challenging due to the time-consuming nature of current microbiological and histopathological methods.
• Delayed diagnosis of bone infections can lead to prolonged treatment and poorer patient outcomes.

Purpose of the Study:

• To evaluate handheld near-infrared (NIR) spectroscopy for detecting spectral alterations in ex vivo human bone inoculated with Staphylococcus species.
• To assess the feasibility of NIR spectroscopy as a rapid, non-destructive diagnostic tool for bone infections.

Main Methods:

• An ex vivo model using human trabecular bone specimens inoculated with Staphylococcus aureus and Staphylococcus epidermidis was created.
• Spectra were acquired using a handheld NIR spectroscopy platform.
• Data preprocessing involved standard normal variate transformation, smoothing, or first-derivative transformation, followed by principal component analysis and linear discriminant analysis.

Main Results:

• Principal component analysis showed partial clustering of inoculated versus uninoculated bone, especially after derivative preprocessing, though overlap persisted.
• Linear discriminant analysis achieved better discrimination between inoculated and uninoculated bone than between the two bacterial species.
• Classification performance was sensitive to preprocessing techniques and validation strategies, necessitating donor-wise partitioning to mitigate information leakage.

Conclusions:

• Handheld NIR spectroscopy demonstrates potential for providing rapid, non-destructive insights into spectral changes associated with bacterial inoculation in ex vivo bone.
• Further research, including clinical validation and testing with diverse pathogens and confounders, is essential before clinical application for diagnosing bone infections.