Artificial Intelligence in Clinical Practice How AI Technologies Impact Medical Research and Clinics: Chapter 45 - Artificial intelligence and global health

Elsevier, Artificial Intelligence in Clinical Practice, How AI Technologies Impact Medical Research and Clinics 2024, Pages 395-399
Jay Vietas

The use of modern machine learning and natural language processing is an emerging area in the prevention and detection of adverse drug events at the point of care, which has the potential to substantially improve medication-related decision-making compared with rule-based tools. This chapter focuses on AI-based algorithms and tools that can be used to inform clinical decision-making for prevention or early detection of adverse drug events at the point of care. This overview addresses three types of AI: rule-based models which are currently used in practice; the next generation of more complex, modern machine learning; and natural language processing-based tools that are under development and in testing. Examples are given of the current rules-based systems that are widely used to support medication-related decision-making are summarized; these use clinical guidelines, patient-specific information such as comorbidities or genomic profiles, and known contraindications from pharmacological databases. The next generation of AI tools that use Machine Learning and Natural Language Processing (NLP) to inform medication-related decision-making are reviewed and there is a brief introduction of the currently limited scope of NLP being applied to extract information from free-text clinical notes in the electronic health record. The current challenges that are limiting the implementation of AI within this field are identified and suggestions for the future direction of AI that will affect clinical decision-making and patient outcomes are given.