Artificial intelligence across organ-on-a-chip workflows: current implementations and future approaches

Organ-on-a-Chip (OoC) systems use microfluidic and tissue engineering techniques to recreate key features of organ microenvironments. They have become a common platform for disease modeling and drug screening. As OoC technologies advance, they generate larger and more diverse datasets. Traditional manual analysis and endpoint measurements are therefore no longer sufficient to capture dynamic processes or long-term responses. Artificial intelligence (AI) enables efficient data analysis and supports functional prediction. This review focuses on how AI is used in OoC experiments, including data acquisition during experiments, post-experimental data analysis, and application-oriented predictive tasks. Current studies mainly apply AI to support real-time sensing and monitoring, computational phenotyping and behavioral analysis, and drug response evaluation. We further discuss several AI approaches that have seen limited use in OoC but may offer new opportunities. Realtime monitoring and intelligent control could improve data acquisition. Multimodal data fusion and deep learning-based prediction may enhance data analysis. At the application level, these methods may further support pharmacology and toxicology studies. These developments indicate a move toward data-driven OoC platforms with improved reliability for precision medicine and drug development.