Artificial Intelligence in Asthma: Current Status, Opportunities, and Pitfalls

Area of Science:

• Computational biology
• Respiratory medicine
• Artificial Intelligence

Background:

• Asthma presents diagnostic and management challenges due to its complexity, diverse phenotypes/endotypes, and common treatment adherence issues.
• Despite available treatments, effective asthma control remains difficult for many patients.
• Advances in AI and ML offer new tools for improving medical decisions and patient outcomes.

Purpose of the Study:

• To review recent literature (last 3 years) on the application of AI and ML in asthma.
• To map current knowledge on AI/ML in asthma diagnosis, risk assessment, classification, and management.
• To identify future research directions for AI/ML in precision asthma care.

Main Methods:

• Narrative review of scientific literature published in the last three years.
• Focus on studies integrating AI/ML techniques into various aspects of asthma care.
• Analysis of applications in risk identification, screening, diagnosis, classification, exacerbation prediction, and guided treatment.

Main Results:

• AI/ML show significant promise in enhancing asthma diagnosis, classification, and risk prediction by analyzing large datasets.
• Applications span early diagnosis, personalized treatment strategies, and exacerbation prevention.
• AI/ML tools are increasingly integrated into asthma management, supporting precision medicine.

Conclusions:

• AI and ML are valuable instruments for improving accuracy in asthma diagnosis, classification, and risk assessment.
• These technologies have the potential to personalize asthma treatment and prevent exacerbations.
• AI will augment, not replace, clinicians, optimizing medical decisions and practices in asthma care.