Highmore Callum, Hanrahan Niall, Cook Yoshiki, Pritchard Ysanne, Lister Adam, Cooper Kirsty, Devitt George, Munro Alasdair P S, Faust Saul N, Mahajan Sumeet, Webb Jeremy S
School of Biological Sciences, Faculty of Environmental and Life Sciences, University of Southampton, SO17 1BJ, Southampton, UK.
National Biofilms Innovation Centre (NBIC) and Institute for Life Sciences, University of Southampton, SO17 1BJ, Southampton, UK.
NPJ Antimicrob Resist. 2025 Aug 25;3(1):74. doi: 10.1038/s44259-025-00141-z.
Antimicrobial resistance (AMR) poses a global healthcare challenge, where overprescription of antibiotics contributes to its prevalence. We have developed a rapid multi-excitation Raman spectroscopy methodology (MX-Raman) that outperforms conventional Raman spectroscopy and enhances specificity. A support vector machine (SVM) model was used to identify 20 clinical isolates of Pseudomonas aeruginosa with an accuracy of 93% using MX-Raman. Antibiotic sensitivity profiles for tobramycin, ceftazidime, ciprofloxacin, and imipenem were generated for the bacterial strains and compared with their Raman spectral signatures using MX-Raman. The 20 clinical strains were distinguished according to AMR profiles. Nine models were assessed for AMR classification performance, and SVM performed best, classifying AMR profiles of each strain with 91-96% accuracy. These data provide the basis for a new rapid clinical diagnostic platform that could screen for bacterial infection and recommend effective antibiotic treatment ahead of confirmation by conventional techniques, improving clinical outcomes and reducing the spread of AMR.
抗菌药物耐药性（AMR）对全球医疗保健构成挑战，抗生素的过度处方是其普遍存在的原因之一。我们开发了一种快速多激发拉曼光谱方法（MX-拉曼），该方法优于传统拉曼光谱并提高了特异性。使用支持向量机（SVM）模型，通过MX-拉曼识别20株铜绿假单胞菌临床分离株，准确率达93%。利用MX-拉曼为这些细菌菌株生成了妥布霉素、头孢他啶、环丙沙星和亚胺培南的抗生素敏感性谱，并将其与拉曼光谱特征进行比较。根据AMR谱对这20株临床菌株进行了区分。评估了9种模型的AMR分类性能，其中SVM表现最佳，对每种菌株AMR谱的分类准确率为91%-96%。这些数据为一个新的快速临床诊断平台奠定了基础，该平台可在通过传统技术确认之前筛查细菌感染并推荐有效的抗生素治疗，改善临床结果并减少AMR的传播。
