| | 講 題 | Identification of Imipenem-nonsusceptible Acinetobacter nosocomialis Isolates by Matrix-Assisted Laser Desorption Ionization-Time of Flight Mass Spectrometry | | | 講 者 | 國立中央大學 Chung-Yu Chien, Ming-Yi Peng, Jhen-Ting Liou, Chia-Ru Chung, Jorng-Tzong Horng | | | 日 期 | 2019/09/27 | 長 度 | | 人 氣 | 62 次 | | | Acinetobacter nosocomialis (A. nosocomialis) is one of the major pathogenic bacteria producing nosocomial infections. The presence of A. nosocomialis in hospital is detrimental to patients and to the hospital management. Rapid identification of imipenem-nonsusceptible A. nosocomialis can provide instant information for administration of antibiotics. Matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS) has recently become a widely used tool for identifying bacterial species. In this study, we constructed a classification model for identifying imipenemnonsusceptible A. nosocomialis based on based on a largescale MALDI-TOF MS dataset consisting of 1085 imipenem-nonsusceptible A. nosocomialis and 1047 imipenem-susceptible A. nosocomialis clinical isolates. The mass-to-charge ratio (m/z) values were mainly adopted to build the machine learning (ML) models, including random forest(RF), decision tree, and logistic regression. Finally, the performance of RF were the best among the other models. As a result, the RF-based prediction model for identifying imipenem-nonsusceptible A. nosocomialis can aid clinicians in administering the appropriate antibiotics.
| | | 提 供 | TANET台灣網際網路研討會-TANET2019 | | |
|
