Discriminating bacterial types in oral and maxillofacial space infections (OMSIs) via smelling diagnosis.

Abstract

Objectives: To explore the feasibility of using gas chromatography-mass spectrometry (GC-MS) in the diagnosis of bacterial types of oral and maxillofacial space infections (OMSIs).

Methods: Staphylococcus aureus was cultured to produce a full spectrum of volatile organic compounds (VOCs). These VOCs were collected and identified using VOC collectors and GC-MS. A bacteria Probability-based VOC Algorithm (BPVA) was developed to assess the likelihood of infection in vivo utilizing a rat model. VOCs from 19 clinical OMSIs abscess samples were analyzed to calculate the probability of SA infection, with results validated against traditional culture results.

Results: Using fold change (FC), p-values (P), and variable importance in projection (VIP) values, we identified 318 characteristic target compounds associated with Staphylococcus aureus. After fitting these compounds to the rat infection model, 189 compounds were selected, establishing a BPVA threshold of ≥ 50, with a 95% probability of consistency. Four patients exhibited an SA infection probability exceeding 70%, whereas the other patients fell below this threshold. Clinical results confirmed SA infections in these four patients, with detection times reduced to approximately 24 h, significantly shorter than the typical 3-5 days.

Conclusions: The combination of GC‒MS and BPVA significantly accelerated the diagnosis of bacterial types in OMSIs, demonstrating strong agreement with traditional culture results, highlighting a new method to accurately and rapidly diagnose maxillofacial infections.

Clinical relevance: Rapid identification of bacterial types is essential for treating OMSIs. However, current bacterial culture methods face challenges, such as limited accuracy and time-consuming processes.