Navigation auf


Digital Society Initiative

Health Data Repository (HDR)

The aim of the project is to establish an infrastructure needed to store and aggregate existing high‐quality musculoskeletal imaging datasets to create and share corresponding data-driven models. The HDR Digital Platform will connect orthopedic medicine and data science addressing structured data storage, accessibility, and findability in musculoskeletal research.

Artificial intelligence and machine learning will significantly influence modern clinical workflows but some questions for the clinical implementation of these methods need to be addressed. Specifically, current efforts trying to accelerate the digitalization of medical research are focusing on routine clinical and molecular data, neglecting musculoskeletal (MSK) data about orthopedic treatment. We are working towards a central open‐science, curated repository focusing on semantically structured anatomical imaging datasets contributing actively to the development of trustworthy ML models and digital solutions supporting researchers, physicians, and ultimately patients, while maintaining a high level of data privacy. The HDR Platform aims to offer an accessible level to anonymous datasets and data‐driven models, where on one hand the testing of novel methods on highly realistic clinical data will be possible, and on the other hand it will support the methodological development in the digital health community thanks to open science.

Thanks to the HDR platform, a novel reference framework supporting scientific output in data‐driven orthopedic medicine will be established. We focus on the development of Digital Twins which may generally be regarded as a combination of statistical shape models (SSMs), encompassing radiomics and structural features, and become increasingly attractive also in the planning and execution of orthopedic surgery. The interface of the HDR Digital Platform to real‐world data in the clinical setting, will allow investigation into trustworthiness and explainability of ML models generated from the datasets.

The embedment of the HDR Platform as collaborative project between the Balgrist University Hospital and the Digital Society Initiative allows to leverage the collaborative ecosystem including the OR-X network, the ZHAW Digital Health Lab and the LOOP network to support the establishment of a standardized database of orthopedic MSK data.

Here you can find a short interview about the project.


Project duration: 01.01.2023 – 31.12.2024




Project Team

Dr. Sebastiano Caprara

Dr. Sebastiano Caprara received his PhD degree at the ETH Zurich with focus on machine learning and predictive models providing solutions for preoperative planning of spinal fusion surgery. During his doctoral studies, he worked for a startup company in close collaboration with the Balgrist University Hospital gaining experience in translational research projects. He is currently leading the Health Data Repository project at the Balgrist University Hospital aiming at the development of a flexible digital infrastructure supporting data-driven clinical research.