An international team of scientists has utilised Artificial Intelligence (AI) to monitor levels of antibiotic resistant bacteria around the world. The findings may help develop targeted strategies to combat the global threat of antimicrobial resistance (AMR).
The team, which included Professor Robert Beardmore of the University of Exeter’s Living Systems Institute, employed mathematical modelling to examine patterns of antibiotic resistant bacteria from 6.5 million data points around the world. The researchers analysed datasets from health agencies and medical companies to develop the most comprehensive picture of the AMR landscape to date.
The study, titled “Seeking patterns of antibiotic resistance in ATLAS, an open, raw MIC database with patient metadata“, was published in Nature Communications.
Combatting AMR with AI
The team utilised a dataset from Pfizer called ATLAS, which has been collecting data for 20 years and is regarded as the most medically complete dataset. Using ATLAS, the researchers evaluated information to identify data gaps and predict future antibiotic resistant bacteria increases.
The results identified resistance not previously seen before and a lack of data in Africa. They explained that the innovative AI method will enhance knowledge of what resistance is exacerbating and where it is located.
Professor Beardmore, the lead author of the study, commented: “AI is a box of tricks that could help solve the antimicrobial resistance problem, but national health agencies need to publish much more data for that to happen.
“Resistance does seem to be on the increase, but even if it were to come down because of successful health policy changes or new medical technologies, missing data make those reductions hard to spot.”
Threats posed by antibiotic resistant bacteria
Antimicrobial resistance is one of the greatest threats to human health in the 21st century, with estimates suggesting that AMR was attributed to five million deaths in 2019. AMR occurs naturally through the overuse of antibiotics, causing the bacteria they are employed to destroy to become resistant to them. The increase in antibiotic resistant bacteria means that illnesses such as sepsis, pneumonia, and tuberculosis are increasingly challenging to treat.
A crucial factor for reducing antibiotic resistant bacteria is identifying patterns of where outbreaks may occur and what medications are impacted. The new AI technique may be an essential tool for this, determining resistance patterns from millions of global data points.
Pablo Catalan, a data scientist from Madrid, said: “Our work on ATLAS shows resistance in patients is a dynamic process that contains lots of patterns, and the more data we have, the better we’ll understand those patterns and understand when our actions are pulling resistance downwards, or making things worse.”
Jon Iredell, a collaborating clinician, added: “When a bacterium is sampled from one of my patients, that is a chance to increase the amount of data we have on patterns of antibiotic resistant bacteria that we are not, as yet, taking.”
