Health UK (Commonwealth Union) – In observance of World Antimicrobial Awareness Week, researchers affiliated with the Oxford Martin Programme on Antimicrobial Resistance Testing at the University of Oxford have unveiled significant progress in developing a novel and expeditious antimicrobial susceptibility test. This innovative test has the capability to deliver results in as little as 30 minutes, presenting a substantial improvement over current gold-standard methodology.
Their findings, detailed in a publication in Communications Biology, showcase the utilization of a fusion of fluorescence microscopy and artificial intelligence (AI) for the detection of antimicrobial resistance (AMR). The methodology involves the training of deep-learning models to analyze images of bacterial cells, identifying structural alterations indicative of responses to antibiotic treatment. Across various antibiotics, the approach demonstrated noteworthy efficacy, achieving a minimum of 80% accuracy on a per-cell basis.
The researchers posit that their model holds the potential to ascertain the resistance of cells in clinical samples to a diverse range of antibiotics in the future. This advancement marks a significant stride in the quest for more rapid and comprehensive antimicrobial susceptibility testing.
Co-author for the paper Achillefs Kapanidis, Professor of Biological Physics who is also Director of the Oxford Martin Programme on Antimicrobial Resistance Testing, explained “Antibiotics that stop the growth of bacterial cells also change how cells look under a microscope, and affect cellular structures such as the bacterial chromosome. Our AI-based approach detects such changes reliably and rapidly. Equally, if a cell is resistant, the changes we selected are absent, and this forms the basis for detecting antibiotic resistance.”
The researchers conducted tests using their approach on a variety of clinical isolates of E. coli, each exhibiting different degrees of resistance to the antibiotic ciprofloxacin. Demonstrating remarkable reliability, the deep-learning models identified antibiotic resistance with a speed surpassing established state-of-the-art clinical methods, considered the gold standard, by at least 10 times.
The team aims to further refine their method, enhancing its speed and scalability for clinical application, while also adapting it for various bacteria and antibiotics.
According to the Global Research on Antimicrobial Resistance (GRAM) Project—a collaboration involving the University—approximately 1.3 million individuals succumbed to antimicrobial resistance (AMR) in 2019.
Current testing techniques rely on cultivating bacterial colonies in the presence of antibiotics. However, these methods are sluggish, often necessitating several days to assess the resistance levels of bacteria to various antibiotics.
This poses a challenge when dealing with patients facing potentially life-threatening infections like sepsis that demand urgent treatment. In such cases, physicians are often compelled to prescribe antibiotics based on clinical experience or a combination of antibiotics proven effective against a range of bacterial infections. However, the use of ineffective antibiotics can exacerbate the patient’s condition, leading to the need for additional antibiotic treatments. A potential consequence of this cycle is an escalation of AMR in the community.
The researchers highlight that further development of their rapid method could pave the way for targeted antibiotic treatments. This, in turn, could reduce treatment durations, minimize side effects, and ultimately contribute to slowing down the surge of antimicrobial resistance.
Co-author of the paper Aleksander Zagajewski, a doctoral student at the University of Oxford, Department of Physics, added “Time is beginning to run out for our antibiotic arsenal; we are hoping our novel diagnostics will pave the way for a new generation of precision treatments for the most sick patients.”
The research paper titled ‘Deep learning and single-cell phenotyping for rapid antimicrobial susceptibility detection in Escherichia coli’ has appeared in Communications Biology.
The investigation drew contributions from various departments within the University, encompassing the Department of Physics, Nuffield Department of Medicine, Sir William Dunn School of Pathology, and Nuffield Department of Women’s Reproductive Health. Additionally, researchers from the Department of Microbiology and Infectious Diseases at the Oxford University Hospitals NHS Foundation Trust played a vital role in the study.
