THE antibiotic ciprofloxacin should stop E. coli bacteria in their tracks, yet the strain I am looking at in a microscope image is resisting the drug鈥檚 effect. The signs of this are subtle, but I have learned to spot them 鈥� and so can you 鈥� by participating in the .
The researchers behind this endeavour have an : to develop a test that uses artificial intelligence to rapidly predict how bacteria are likely to respond to commonly available antibiotics. Using thousands of microscope images of antibiotic-resistant and antibiotic-sensitive bacteria in the presence of different drugs, at the University of Oxford and her colleagues have built a machine-learning tool that can detect signs of antibiotic resistance when presented with similar images. But they need our help to refine it.
As a newly certified Infection Inspector, you will be shown microscope images of bacteria exposed to various antibiotics and be asked to indicate whether the bacteria you are viewing appear sensitive to or resistant to the drug in question. You can access the project online via the citizen science platform, where there is a tutorial to teach you what to look for.
Advertisement
In the case of my hardy E. coli cell, what catches my attention is the bacterium鈥檚 DNA, which shows up as fluorescent green. It appears in several distinct clumps, spread throughout the rod-shaped cell. This indicates the bacterium has made copies of its DNA and is poised to replicate itself 鈥� a process that should have been inhibited by the ciprofloxacin. I label it as resistant and click 鈥渄one鈥�.
I am one of hundreds of volunteers who have participated in the project since it launched in February. Farrar hopes our classifications will help develop a rapid diagnostic test that, within an hour, can identify the bacteria causing a particular infection and predict their sensitivity to common antibiotics. This would enable healthcare workers to treat infections promptly and in a targeted way, improving patient outcomes and slowing the rise of antibiotic resistance.
Such resistance is a major threat to public health. A 2022 analysis more than 1.2 million people were killed by antibiotic-resistant infections in 2019. That is to rise to as many as 10 million by 2050.
Current rapid tests can鈥檛 directly predict how bacteria will respond to an antibiotic, says Farrar. To get such information usually requires measuring the growth of cultured bacterial samples 鈥� typically taking at least a day to yield results. A new generation of rapid tests that can do this is needed. 鈥淸Such] tests for bacteria will help reduce the spread of antibiotic resistance because they will help clinicians prescribe the most appropriate, targeted antibiotics earlier in the course of an infection,鈥� says Farrar.
What you need
Access to Infection Inspection via
For other projects visit newscientist.com/maker.
