Detection of carbapenem-resistant Klebsiella pneumoniae on the basis of matrix-assisted laser desorption ionization time-of-flight mass spectrometry by using supervised machine learning approach

• Published in: PloS one Vol. 15; no. 2; p. e0228459
• Main Authors: Huang, Tsi-Shu, Lee, Susan Shin-Jung, Lee, Chia-Chien, Chang, Fu-Chuen
• Format: Journal Article
• Language: English
• Published: United States Public Library of Science 06.02.2020; Public Library of Science (PLoS)
• Subjects: Algorithms; Analysis; Anti-Bacterial Agents - therapeutic use; Antibiotics; Antimicrobial agents; Bacteria; Biology and Life Sciences; Carbapenem-Resistant Enterobacteriaceae - drug effects; Carbapenem-Resistant Enterobacteriaceae - genetics; Carbapenem-Resistant Enterobacteriaceae - isolation & purification; Carbapenems - therapeutic use; Classification; Clinical microbiology; Computer and Information Sciences; Cross Infection - diagnosis; Cross Infection - drug therapy; Cross Infection - microbiology; Data mining; Desorption; DNA, Bacterial - analysis; DNA, Bacterial - isolation & purification; Drug resistance; Enzymes; Health aspects; Hospitals; Humans; Identification methods; Imipenem; Infection; Ionization; Ions; Klebsiella; Klebsiella Infections - diagnosis; Klebsiella Infections - drug therapy; Klebsiella Infections - microbiology; Klebsiella pneumoniae; Klebsiella pneumoniae - genetics; Klebsiella pneumoniae - isolation & purification; Lasers; Learning algorithms; Machine learning; Mass spectrometry; Mass spectroscopy; Medical laboratories; Medicine; Medicine and Health Sciences; Meropenem; Methicillin; Microbial Sensitivity Tests; Microbiology; Microorganisms; Pathology; Physical Sciences; Pneumonia; Predictive Value of Tests; Random Amplified Polymorphic DNA Technique; Research and Analysis Methods; Ribosomal DNA; Scientific imaging; Sodium; Spectrometry, Mass, Matrix-Assisted Laser Desorption-Ionization - methods; Spectroscopy; Staphylococcus infections; Supervised Machine Learning; Support vector machines; Teaching methods; Time; United States > US; Taiwan
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0228459

Carbapenem-resistant Klebsiella pneumoniae (CRKP) is emerging as a significant pathogen causing healthcare-associated infections. Matrix-assisted laser desorption/ionisation mass spectrometry time-of-flight mass spectrometry (MALDI-TOF MS) is used by clinical microbiology laboratories to address the need for rapid, cost-effective and accurate identification of microorganisms. We evaluated application of machine learning methods for differentiation of drug resistant bacteria from susceptible ones directly using the profile spectra of whole cells MALDI-TOF MS in 46 CRKP and 49 CSKP isolates. We developed a two-step strategy for data preprocessing consisting of peak matching and a feature selection step before supervised machine learning analysis. Subsequently, five machine learning algorithms were used for classification. Random forest (RF) outperformed other four algorithms. Using RF algorithm, we correctly identified 93% of the CRKP and 100% of the CSKP isolates with an overall classification accuracy rate of 97% when 80 peaks were selected as input features. We conclude that CRKPs can be differentiated from CSKPs through RF analysis. We used direct colony method, and only one spectrum for an isolate for analysis, without modification of current protocol. This allows the technique to be easily incorporated into clinical practice in the future.