Leveraging artificial intelligence for the management of preschool wheeze: A narrative review

Management of preschool wheeze is notoriously challenging given heterogeneous clinical trajectories and underlying biological mechanisms dictating therapeutic response. Data‐driven approaches have highlighted the value of identifying individual wheeze phenotypes and underlying biomarkers to support a personalized management approach; however, these advancements have yet to be translated into clinical management. Here, we discuss key opportunities for Artificial Intelligence and Machine Learning to support personalized approaches to wheeze management through vast pattern‐recognition capabilities. Advancements in the development of tools for objective symptom evaluation, remote symptom monitoring, and prediction of clinical trajectories are summarized. Key considerations for the responsible and successful deployment of such promising technologies in real‐world clinical settings are emphasized, including prevention of algorithmic biases, promotion of prediction transparency, and establishing standards for patient data privacy and equitable access to novel technologies.