From data governance to intelligent applications: practice and reflection on building the medical “dry laboratory ”

This study aims to develop a medical “dry laboratory” model that treats healthcare data as the primary research subject and employs computational science and AI as core methodologies. Based on the practice of a large tertiary hospital,we present a construction framework featuring a full-cycle model spanning data cultivation, governance, value extraction, and ecological feedback—transforming medical data from a static resource into a living asset. Innovatively, the “dry laboratory” was integrated into the residency training rotation to systematically cultivate data literacy. Practice has shown that the framework establishes the “dry laboratory” as a digital research entity, creating a closed-loop system that integrates clinical practice with data analytics, significantly enhancing the usability of multimodal data and enabling scalable training in data competence. In conclusion, the medical “dry laboratory” represents a new digital research paradigm distinct from traditional wet laboratories. Its implementation provides a reference framework and introduces an integrated “dry-wet” research model combining wet-lab experiments with data cultivation, demonstrating significant value for enhancing medical innovation, cultivating interdisciplinary talent, and advancing smart healthcare.