Urinary tract infection with mixed microorganisms may lead to false-positive resistance detection.Current antimicrobial susceptibility testing(AST)performed in clinical laboratories is based on bacterial culture and t...Urinary tract infection with mixed microorganisms may lead to false-positive resistance detection.Current antimicrobial susceptibility testing(AST)performed in clinical laboratories is based on bacterial culture and takes a long time for mixed bacterial infections.Here,we propose a machine learning-based single-cell metabolism inactivation concentration(ML-MIC)model to achieve rapid AST for mixed bacterial infections.Using E.coli and S.aureus as a demonstration of mixed bacteria,we performed feature extraction and multi-feature analysis on stimulated Raman scattering(SRS)images of bacteria with the ML-MIC model to determine the subtypes and AST of the mixed bacteria.Furthermore,we assessed the AST of mixed bacteria in urine and obtained single-cell metabolism inactivation concentration in only 3 h.Collectively,we demonstrated that SRS imaging of bacterial metabolism can be extended to mixed bacterial infection cases for rapid AST.展开更多
Rapid antimicrobial susceptibility testing(AST)is urgently needed to slow down the emergence of antibioticresistant bacteria and treat infections with correct antibiotics.Stimulated Raman scattering(SRS)microscopy is ...Rapid antimicrobial susceptibility testing(AST)is urgently needed to slow down the emergence of antibioticresistant bacteria and treat infections with correct antibiotics.Stimulated Raman scattering(SRS)microscopy is a technique that enables rapid chemical-bond imaging with sub-cellular resolution.It can obtain the AST results with a single bacterium resolution.Although the SRS imaging assay is relatively fast,taking less than 2 h,the calculation of single-cell metabolism inactivation concentration(SC-MIC)is performed manually and takes a long time.The bottleneck tasks that hinder the SC-MIC throughput include bacterial segmentation and intensity thresholding.To address these issues,we devised a hybrid algorithm to segment single bacteria from SRS images with automatic thresholding.Our proposed method comprises a U-Net convolutional neural network(CNN),DropBlock,and secondary segmentation post-processing.Our results show that SC-MIC calculation can be accomplished within 1 min and more accurate segmentation results using deep learning-based bacterial segmentation method,which is essential for its clinical applications.展开更多
基金the National Natural Science Foundation of China(81901790)the Key R&D program of Ministry of Science and Technology(2020YFC2005405)Beijing Natural Science Foundation(No.7224367 to X.Chen).
文摘Urinary tract infection with mixed microorganisms may lead to false-positive resistance detection.Current antimicrobial susceptibility testing(AST)performed in clinical laboratories is based on bacterial culture and takes a long time for mixed bacterial infections.Here,we propose a machine learning-based single-cell metabolism inactivation concentration(ML-MIC)model to achieve rapid AST for mixed bacterial infections.Using E.coli and S.aureus as a demonstration of mixed bacteria,we performed feature extraction and multi-feature analysis on stimulated Raman scattering(SRS)images of bacteria with the ML-MIC model to determine the subtypes and AST of the mixed bacteria.Furthermore,we assessed the AST of mixed bacteria in urine and obtained single-cell metabolism inactivation concentration in only 3 h.Collectively,we demonstrated that SRS imaging of bacterial metabolism can be extended to mixed bacterial infection cases for rapid AST.
基金This research was supported by the Natural Science Foundation of Beijing(7204274)the National Natural Science Foundation of China(81901790).
文摘Rapid antimicrobial susceptibility testing(AST)is urgently needed to slow down the emergence of antibioticresistant bacteria and treat infections with correct antibiotics.Stimulated Raman scattering(SRS)microscopy is a technique that enables rapid chemical-bond imaging with sub-cellular resolution.It can obtain the AST results with a single bacterium resolution.Although the SRS imaging assay is relatively fast,taking less than 2 h,the calculation of single-cell metabolism inactivation concentration(SC-MIC)is performed manually and takes a long time.The bottleneck tasks that hinder the SC-MIC throughput include bacterial segmentation and intensity thresholding.To address these issues,we devised a hybrid algorithm to segment single bacteria from SRS images with automatic thresholding.Our proposed method comprises a U-Net convolutional neural network(CNN),DropBlock,and secondary segmentation post-processing.Our results show that SC-MIC calculation can be accomplished within 1 min and more accurate segmentation results using deep learning-based bacterial segmentation method,which is essential for its clinical applications.