Using FTIR and Machine Learning to Detect MRSA 

Antimicrobial resistance (AMR) poses a significant threat to global health, with Methicillin-resistant Staphylococcus aureus (MRSA) contributing substantially to morbidity. Rapid identification of MRSA versus Methicillin-susceptible S. aureus (MSSA) is critical for timely and appropriate antibiotic use. This presentation introduces applications of machine learning, focusing on Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA), as tools for analyzing Fourier Transform Infrared (FTIR) spectroscopy data to efficiently distinguish MRSA from MSSA. By analyzing the growth patterns and spectral data of SA6538 (MSSA) and SA43300 (MRSA) under antibiotic stress, this work highlights novel approaches to spectral loading curve analysis, revealing biochemical changes over time and across strains. In addition we will demonstrate how LDA applied to PCA-reduced datasets achieves high classification accuracy for resistant strains in less than one hour. This study not only showcases the potential of a rapid diagnostic method for differentiating antibacterially perturbed MRSA and MSSA but also introduces innovative strategies for spectral analysis, offering deeper insights into chemical changes in response to antibiotic exposure. These findings represent a significant step toward understanding AMR and improving clinical outcomes in the fight against its spread.