One early adopter's tips for AI deployment success

Dr. Luis Ahumada, the Director of Health Data Science and Analytics at John Hopkins All Children's Hospital, will discuss his experiences with AI and machine learning at the HIMSS AI in Healthcare Forum in San Diego. As an early adopter of AI, he will share real-world examples of how Johns Hopkins is utilizing AI-enabled tools to enhance care-delivery processes, such as automating clinical diagnostics and improving efficiency.

Dr. Ahumada emphasizes the importance of understanding the main benefits of AI, especially considering limited resources and the high cost of AI implementation. He sees value in two types of AI: Large Language Models (LLMs) and traditional machine learning, with a focus on creating models for prediction and high-risk calculators. Johns Hopkins has concentrated on clinical documentation as a key use case, addressing inefficiencies in the system.

Data integrity is crucial for AI applications, and Dr. Ahumada stresses the need for improved data collection to enhance machine learning capabilities. Challenges include missing data, validation issues, and the difficulty of assembling large datasets. Despite these challenges, Johns Hopkins is actively pursuing various generative AI and machine learning use cases.

Dr. Ahumada expresses concerns about the cost of AI implementation, noting that while large health systems are leveraging generative AI, smaller ones face limitations due to financial constraints. He advocates for making AI tools accessible to health systems of all sizes to improve clinical and administrative processes. Overall, he sees AI as an ongoing project with significant benefits but acknowledges the associated costs and challenges.