AI-assisted MALDI-TOF MS for identifying carbapenem resistance in clinical <i>Acinetobacter baumannii</i> isolates.
AI-assisted MALDI-TOF MS achieved an accuracy of 96.36% in predicting carbapenem resistance in Acinetobacter baumannii isolates, which may enhance clinical decision-making.
Where it sits
this study against the rest of the evuzamitide corpusSummary and findings
This study evaluated the use of AI-assisted MALDI-TOF MS to predict carbapenem resistance in Acinetobacter baumannii isolates. A total of 191 clinical and surveillance isolates were analyzed, achieving an accuracy of 96.36% in distinguishing carbapenem-resistant from susceptible strains. The findings suggest potential molecular correlates of resistance acquisition.
Abstract
Carbapenem-resistant <i>Acinetobacter baumannii</i> (CRAB) is one of the most critical public health threats worldwide due to its high infection rates, substantial mortality, and limited therapeutic choices. As CRAB infections are frequently multidrug-resistant, rapid and accurate determination of carbapenem susceptibility is essential for appropriate therapeutic decision-making. We established an integrated framework combining matrix-assisted laser desorption ionization-time-of-flight mass spectrometry (MALDI-TOF MS) with artificial intelligence (AI) to enable rapid prediction of carbapenem resistance in <i>A. baumannii</i>. A total of 191 clinical and surveillance isolates, including CRAB and carbapenem-susceptible <i>A. baumannii</i> (CSAB), were recovered from hospitalized patients and phenotypically characterized by standard minimum inhibitory concentration testing. MALDI-TOF MS spectra were subsequently acquired, and six AI models were developed and systematically investigated for predictive performance, among which the eXtreme Gradient Boosting (XGBoost) model achieved the highest discriminatory performance, distinguishing CRAB from CSAB with an accuracy of 96.36% and robust overall performance. Feature importance analysis of the XGBoost model revealed that its high predictive performance was driven partially by spectral features associated with horizontal gene transfer-related elements and membrane and transport-associated proteins, providing candidate molecular correlates of carbapenem resistance. Overall, these results show that AI-assisted MALDI-TOF MS enables rapid and accurate prediction of carbapenem resistance in <i>A. baumannii</i> and provides insights into the molecular features associated with resistance acquisition.
Background
Carbapenem-resistant Acinetobacter baumannii (CRAB) poses a significant public health threat due to its high infection rates and limited treatment options. Previous methods for determining carbapenem susceptibility have been time-consuming, necessitating the development of faster diagnostic techniques. This study investigates the integration of AI with MALDI-TOF MS to enhance the speed and accuracy of carbapenem resistance detection.
Methods
The study utilized an integrated framework combining MALDI-TOF MS and AI to analyze 191 clinical and surveillance isolates of A. baumannii. The isolates included both CRAB and carbapenem-susceptible A. baumannii (CSAB). The primary outcome measure was the accuracy of the predictive models developed, particularly the XGBoost model.
Results
The XGBoost model achieved an accuracy of 96.36% in distinguishing CRAB from CSAB. The study does not report on additional statistical measures such as p-values or confidence intervals. Feature importance analysis indicated that spectral features related to gene transfer and membrane proteins contributed to the model's predictive performance.
Interpretation
The high accuracy of the XGBoost model suggests that AI-assisted MALDI-TOF MS could be a valuable tool for rapid identification of carbapenem resistance. However, the clinical significance of this finding remains uncertain, as the study does not address how these results translate into patient outcomes or treatment decisions. Limitations such as the lack of clinical outcome data and potential biases in model training could confound the conclusions.
Key findings
- Accuracy of 96.36% for the XGBoost model in distinguishing CRAB from CSAB.
- 191 clinical and surveillance isolates analyzed.
- Not reported in abstract.
Limitations
- No clinical outcome data reported.
- Potential biases in AI model training.
- Single-site study, limiting generalizability.
- Not reported in abstract.