AI-based mobile application to fight antibiotic resistance

Area of Science:

• Medical Microbiology
• Artificial Intelligence in Healthcare
• Global Health

Background:

• Antimicrobial resistance (AMR) is a critical global health challenge.
• Antibiotic misuse accelerates the development of AMR.
• Traditional disk diffusion antibiotic susceptibility testing (AST) suffers from operator variability and complex interpretation.

Purpose of the Study:

• To develop and evaluate an AI-based, offline smartphone application for automated antibiogram analysis.
• To provide an accessible and reliable AST solution for resource-limited settings.

Main Methods:

• Development of an AI-driven smartphone application for antibiogram image capture and analysis.
• Implementation of a user-friendly graphical interface and an embedded expert system for data validation and interpretation.
• Validation of the application's performance against a hospital-standard automatic system and manual measurements.

Main Results:

• The AI application achieved 90% agreement with a hospital-standard automatic system and 98% agreement with manual measurements for susceptibility categorization.
• The system demonstrated reduced inter-operator variability compared to traditional methods.
• The application enables fully automatic measurement and interpretation of AST results on a smartphone.

Conclusions:

• Automatic reading of antibiotic resistance testing is feasible using smartphone technology.
• The developed AI application is suitable for resource-limited settings, potentially improving global access to AST.
• This technology can significantly enhance patient access to timely and accurate diagnostic information for bacterial infections worldwide.