Researchers at the University of Cambridge have demonstrated that artificial intelligence (AI) can be used to identify drug-resistant infections, significantly reducing the time needed for accurate diagnosis. The team illustrated that an algorithm could be trained to correctly identify drug-resistant bacteria from microscopy images alone.
Antimicrobial resistance is an escalating global health issue, rendering many infections difficult to treat and reducing available treatment options. This situation raises the alarming possibility of some infections becoming untreatable in the near future.
One of the main challenges faced by healthcare professionals is the rapid differentiation between organisms treatable with first-line drugs and those resistant to such treatments. Traditional testing methods can take several days, as bacteria must be cultured, tested against various antimicrobial treatments, and analysed by a technician or machine. This delay often results in patients receiving inappropriate medication, leading to more severe outcomes and potentially exacerbating drug resistance.
In research published in Nature Communications, a team led by researchers in Professor Stephen Baker’s Lab at the University of Cambridge developed a machine learning (ML) tool capable of identifying from microscopy images Salmonella typhimurium bacteria resistant to the first-line antibiotic ciprofloxacin, without needing to test the bacteria against the drug.
Salmonella typhimurium causes gastrointestinal illness and severe typhoid-like illness, with symptoms including fever, fatigue, headache, nausea, abdominal pain, and constipation or diarrhoea. In severe cases, it can be life-threatening. Although infections can be treated with antibiotics, the bacteria are increasingly resistant to several antibiotics, complicating treatment.
The researchers used high-resolution microscopy to examine S. typhimurium isolates exposed to increasing concentrations of ciprofloxacin and identified the five most significant imaging features for distinguishing between resistant and susceptible isolates.
They then trained and tested an ML algorithm to recognise these features using imaging data from 16 samples.
The algorithm successfully predicted whether bacteria were susceptible or resistant to ciprofloxacin in each case, without needing to expose the bacteria to the drug. This was possible for isolates cultured for only six hours, compared to the usual 24 hours required to culture a sample in the presence of an antibiotic.
Dr Tuan-Anh Tran, who worked on this research while a PhD student at the University of Oxford and is now based at the University of Cambridge, said: “S. Typhimurium bacteria that are resistant to ciprofloxacin have several notable differences to those still susceptible to the antibiotic. While an expert human operator might be able to identify some of these, on their own they wouldn’t be enough to confidently distinguish resistant and susceptible bacteria.
“The beauty of the machine learning model is that it can identify resistant bacteria based on a few subtle features on microscopy images that human eyes cannot detect.”
To analyse a sample using this approach, it remains necessary to isolate the bacteria from a sample, such as blood, urine, or stool. However, as the bacteria do not need to be tested against ciprofloxacin, the entire process could be reduced from several days to a matter of hours.
While there are practical and cost-effectiveness limitations to this approach, the team believes it demonstrates the potential power of artificial intelligence in combating antimicrobial resistance.
Dr Sushmita Sridhar, who initiated this project while a PhD student in the Department of Medicine at the University of Cambridge and is now a postdoc at the University of New Mexico and Harvard School of Public Health, said: “Given that this approach uses single cell resolution imaging, it isn’t yet a solution that could be readily deployed everywhere. But it shows real promise that by capturing just a few parameters about the shape and structure of the bacteria, it can give us enough information to predict drug resistance with relative ease.”
The team now aims to work with larger collections of bacteria to create a more robust experimental set that could further accelerate the identification process and allow them to identify resistance to ciprofloxacin and other antibiotics in various bacterial species.
Sridhar added: “What would be really important, particularly for a clinical context, would be to be able to take a complex sample – for example blood or urine or sputum – and identify susceptibility and resistance directly from that. That’s a much more complicated problem and one that really hasn’t been solved at all, even in clinical diagnostics in a hospital. If we could find a way of doing this, we could reduce the time taken to identify drug resistance and at a much lower cost. That could be truly transformative.”
The research was funded by Wellcome.
There’s plenty of other editorial on our sister site, Electronic Specifier! Or you can always join in the conversation by commenting below or visiting our LinkedIn page.
