jmm editor's choice: a same-day antimicrobial susceptibility test

posted on june 12, 2020   by 英格兰vs美国谁会赢？

the journal of medical microbiology (jmm) is a journal published by the 英格兰vs美国谁会赢？ , focused on providing comprehensive coverage of medical, dental and veterinary microbiology and infectious diseases, including bacteriology, virology, mycology and parasitology. this month, professor roberto la ragione discusses the paper 'same-day antimicrobial susceptibility test using acoustic-enhanced flow cytometry visualized with supervised machine learning' which was selected as editor's choice for the may issue of jmm. 

"antimicrobial resistance (amr) remains a global challenge and accurate antimicrobial susceptibility testing is essential to help avoid amr and treatment failures. however, many of the current antimicrobial susceptibility testing methods are slow and technically demanding.  this manuscript assessed the feasibility of using novel machine-learning techniques for the detailed analysis of data produced using a flow cytometer-assisted antimicrobial susceptibility test (fast). the study demonstrates that the combination of machine learning with the fast method can generate same-day antimicrobial susceptibility test results, and has the potential to aid timely antimicrobial treatment decisions, stewardship and detection of amr."

same-day antimicrobial susceptibility test using acoustic-enhanced flow cytometry visualized with supervised machine learning

the clock starts ticking at the beginning of a serious bacterial infection. any delay in starting effective antibiotic treatment increases the risk of complications and death. to avoid this, doctors have to guess the best choice of antibiotic and hope the lab tests back them up. antibiotic resistance is the spanner in the works. it’s difficult to predict, and often delays the start of effective treatment, since detection of antibiotic resistance can take several days. that is why delivery of faster, accurate antibiotic susceptibility results is a game changer in severe bacterial infection. earlier antibiotic certainty will save lives, suffering and health service costs. 

our fast method analyses bacteria exposed to antibiotics to produce susceptibility results around a day before conventional lab methods. this level of accurate analysis relies on a massive data output from the analytical machine, an acoustic flow cytometer. our recent paper on the fast method shows how we can use machine learning techniques to speed up raw fast data analysis. this, in turn, produces definitive antibiotic susceptibility results the same day as the earliest indication of bloodstream infection. the combination of flow cytometer analysis and machine learning is a fast route to rapid antibiotic susceptibility.