Abstract
Artificial intelligence (AI) and digital health technologies are increasingly used in the medical field. Despite promises of leading the future of personalized medicine and better clinical outcomes, implementation of AI faces barriers for deployment at scale. We introduce a novel implementation framework that can facilitate digital health designers, developers, patient groups, policymakers, and other stakeholders, to co-create and solve issues throughout the life cycle of designing, developing, deploying, monitoring, and maintaining algorithmic models. This framework targets health systems that integrate multiple machine learning (ML) models with various modalities. This design thinking approach promotes clinical utility beyond model prediction, combining privacy preservation with clinical parameters to establish a reward function for reinforcement learning, ranking competing models. This allows leveraging explainable AI (xAI) methods for clinical interpretability. Governance mechanisms and orchestration platforms can be integrated to monitor and manage models. The proposed framework guides users toward human-centered AI design and developing AI-enhanced health system solutions.