How AI Could Revolutionise The Fight Against Antimicrobial Resistance
Tuberculosis, malaria, and HIV are just three of the world’s most prominent diseases which, according to the World Health Organisation, are beginning to develop strains resistant to drug treatments. This phenomenon - known as antimicrobial resistance (AMR) - occurs when microorganisms (such as bacteria and viruses) evolve the ability to survive antimicrobial treatments through the process of natural selection. In other words, when antimicrobials are overused, inappropriately prescribed, or not used to completion, a small proportion of the target organisms survive due to random mutations and flourish under their newfound lack of competition. As these resistant microbes reproduce and pass on their genetic information, entire generations of antimicrobial-resistant strains arise with the potential to cause rapidly-spreading infections only treatable with a dwindling range of substances. AMR is estimated to cause 25,000 deaths and 2.5 million additional hospital stays annually in Europe alone.

These two images show strains of bacteria spreading into regions of increasingly powerful antibiotics as they evolve antibiotic resistance (Credit: Kishony Lab, Harvard Medical School)
Fortunately, a project funded by the European Institute of Innovation and Technology (EIT) hopes to address these issues with the help of AI. Abtrace, a platform founded by Dr. Umar Naeem Ahmad, Dr. Cristina Correia, and Helder Soares, has designed an algorithm to process a patient’s healthcare notes and recommend the type and dosage of antimicrobial to be administered. By guiding clinicians to making more accurate and beneficial prescriptions for each individual patient, Abtrace hopes to drastically reduce the risk of AMR development.
We are sleepwalking towards a devastating world where antibiotics do not work anymore...we are committed to changing that future starting now.
AI has already been used in healthcare to develop new antibiotics, but Abtrace’s new approach hopes to have a more impactful influence on the fight against AMR. Their algorithm uses a process known as natural language processing to compare patient notes to millions of data points on dosages and success rates in order to find or extrapolate a solution. The application of this technology could be vital to tackling AMR, as Abtrace estimates that 30% of all antibiotic prescriptions are inappropriate.
By the end of May 2020, Abtrace’s platform will receive CE mark certification allowing trading within the European Economic Area. Abtrace has been entitled to up to €2m in funding from the EIT to aid in commercialising this product.
(Thumbnail Source: Asia Pacific Biotech News)