Artificial Intelligence Applications in Antimicrobial Resistance: Comprehensive Review of Predictive Models,...

Area of Science:

• Computational biology
• Infectious disease research
• Medical informatics

Background:

• Antimicrobial resistance (AMR) is a critical global health crisis with significant mortality and economic impacts.
• Traditional methods for combating AMR face limitations in speed and precision.

Purpose of the Study:

• To review recent advancements in artificial intelligence (AI) applications for antimicrobial resistance (AMR).
• To explore AI's role in AMR prediction, antimicrobial discovery, and clinical decision support.

Main Methods:

• Narrative review of peer-reviewed studies from January 2015 to April 2026.
• Analysis of AI models including Deeparg-LS, XGBoost, and vision transformers.
• Evaluation of AI impact on diagnostics and drug discovery.

Main Results:

• AI models demonstrated high accuracy (AUC > 0.90) in AMR prediction.
• AI-driven systems reduced antibiotic mismatches by up to 67%.
• AI accelerated antimicrobial peptide discovery with 76% success and improved susceptibility testing.

Conclusions:

• AI is a pivotal tool in combating AMR, enhancing prediction, discovery, and clinical support.
• Challenges include data issues, computational demands, and ethical considerations.
• Explainable AI and federated learning offer potential solutions for future integration.