Training datasets and AI application toolkits.
Medical AI Infrastructure
Shanghai Shenkang builds a clinically oriented medical big-data training facility.
SenseCare supports a city-level medical AI infrastructure with compute, algorithms, datasets, annotation standards and evaluation tools.
Case Summary
Shanghai Shenkang Hospital Development Center
Shanghai Shenkang Hospital Development Center worked with SenseCare to build a clinically oriented medical big-data training facility. The platform provides one foundational compute infrastructure platform and 24 disease training datasets and AI application toolkits for municipal medical institutions.
Imaging annotation standards and expert consensus references.
CT, pathology, ultrasound and endoscopy.
The project built one foundational compute platform and 24 disease training datasets/toolkits.
The datasets cover 7 common diseases, 7 rare and difficult diseases and 10 major diseases.
Modalities include CT, pathology, ultrasound and endoscopy, covering cancer, cardiovascular, respiratory, metabolic, degenerative, trauma and developmental conditions.
The project established 24 imaging annotation standards and expert consensus references.
The platform provides a one-stop environment with datasets, annotation tools, training engines, algorithm libraries and evaluation tools.
Clinical application as the organizing principle
The Shenkang platform is built from real-world clinical data resources gathered through Shanghai's municipal hospital information-sharing project. The goal is to create representative, authoritative and diverse medical datasets for AI development and validation.
By combining compute infrastructure, desensitized data, annotation tools and evaluation specifications, the project reduces the technical threshold for clinically useful medical AI development.
A shared innovation base for the medical AI ecosystem
The facility is designed for multiple participants: hospitals provide clinical data and scenarios, research groups develop algorithms, companies accelerate product iteration, and regulators gain a more standardized foundation for evaluation.
For physicians and researchers, the platform provides lower-barrier access to model training and validation. For the regional ecosystem, it supports faster medical AI capability building at city scale.
Related Products