APPLICATION OF ARTIFICIAL INTELLIGENCE IN SEARCHING FOR NEW APPROACHES TO FIGHT ANTIMICROBIAL RESISTANCE
SILVIYA GEORGIEVA MIHAYLOVA 1, MOMCHIL KONSTANTINOV LAMBEV 1*, PLAMEN STEFKOV BEKYAROV 1,2, DIMANA DIMITROVA GEORGIEVA 1, MARIYA GAVRAILOVA DANGOVA 1, ANTOANETA ZDRAVKOVA TSVETKOVA ¹
1 Medical college, Medical University of Varna, Varna, Bulgaria
2 Specialized Hospital of Obstetrics and Gynecology for Active Treatment “Prof. Dr. D. Stamatov”, Varna, Bulgaria
Download Full Article PDF
Artificial intelligence (AI) has emerged as a promising tool to support efforts addressing antimicrobial resistance (AMR) by improving diagnostic accuracy, optimizing antimicrobial therapy, and accelerating drug discovery. This review systematically examines studies published between 2013 and 2023 that apply AI-based approaches to combat AMR, with emphasis on clinical decision support and the discovery of new antimicrobial agents, including antimicrobial peptides. A structured literature search was conducted in PubMed, Scopus, and Web of Science following PRISMA guidelines, using predefined search strategies and eligibility criteria. The findings indicate that machine learning and deep learning methods can enhance pathogen identification, support rational antibiotic use, identify synergistic drug combinations, and facilitate the in silico design of novel antimicrobial compounds. In addition, AI-driven models show promising accuracy in predicting antimicrobial peptide activity. Тhe translation of AI-based models into clinical practice is limited by heterogeneous and biased datasets, restricted model interpretability, limited generalisability, and the absence of robust prospective clinical validation studies. Overall, AI offers valuable supportive tools in the fight against AMR; however, further validation and integration into clinical practice are still required.
