A faster way to diagnose antibiotic resistance utilizing AI
As of recently, collaborative research between ETH Zürich, University Hospital Basel, and the University of Basel has been able to demonstrate that machine learning can substantially speed up the detection of antimicrobial resistance compared to using conventional methods. This new approach could help treat severe infections more efficiently in the future.
Antibiotic-resistant bacteria are emerging in countries all over the world – and Switzerland is no exception. Infections caused by multi-drug resistant bacteria lead to a minimum of 300 fatalities in Switzerland every year. Rapid diagnostic testing and the targeted use of antibiotics play an essential role in curbing the spread of these antibiotic-resistant “superbugs”.
However, using conventional detection methods it takes more than two days to determine which antibiotics are suited to treat a particular pathogen because the bacteria from the patient’s sample first have to be cultivated in the lab. Due to this delay, many doctors initially treat serious infections with broad-spectrum antibiotics.
The research team, which forms part of the NCCR AntiResist network, has developed a method that uses mass spectrometry to identify antibiotic resistance in bacteria up to 24 hours earlier.
Find the original research paper here: https://www.nature.com/articles/s41591-021-01619-9