From data silos to smart medicine: How is “dry laboratory” transforming clinical research
image: The closed-loop research model of the medical dry laboratory: clinical data flow into the platform, undergo governance and analysis, produce evidence that re-enters clinical practice, and generate new data for continuous learning. (Image credit: Liu Zifeng, Yang Qintai, Lu Yaxin, Sun Jie, Wang Zhenhe, Rong Limin)
Credit: HIGHER EDUCATION PRESS
Most hospitals sit on a goldmine of electronic health records, medical images, and genomic data—yet the vast majority of these data remain underused, scattered across silos, and lacking the quality needed for research.
Now, a research team from a large tertiary hospital in China has designed and implemented a medical “dry laboratory”—a dedicated digital research platform that treats data as the primary experimental subject and computation as the core research tool. Unlike traditional wet labs that work with cells or animals, the dry lab focuses on data governance, AI modeling, and rapid iteration.
The laboratory operates on a full-cycle closed-loop model:
- Cultivate high-quality data at the source (e.g., structured clinical documentation);
- Govern multimodal data through standardization and quality control;
- Extract value via real-world studies, large language models, and medical imaging AI;
- Re-feed the findings into clinical workflows, generating new data for continuous improvement.
A key innovation is the integration of the dry laboratory into the standardized residency training program. All residents must complete a one-month rotation covering data literacy, research variable definition, and basic AI analytics. This scales up the training of “clinicians who understand data”—a critical workforce for the data-driven era of medicine.
The infrastructure includes a heterogeneous computing cluster (>10 PetaFLOPS FP16) and a unified data platform that supports disease-specific databases, self-service research cohorts, and built-in statistical tools. An institutional three-level governance structure (strategic committee, academic committee, and executive core) ensures strategic alignment, ethical compliance, and operational efficiency.
“The dry laboratory is not just a piece of technology—it is a new research paradigm,” the authors conclude. “We are moving from ‘raising cells and raising mice’ to the integrated model of ‘raising cells, raising mice, and raising data’.”
The work titled “From data governance to intelligent applications: practice and reflection on building the medical ‘dry laboratory’” was published in Journal of New Medicine (published in April 2026).