Unlocking New Horizons in Surgical Data Science

The Data Hub, which records the intraoperative data in the OR-X in real time.

Unlocking New Horizons in Surgical Data Science

Overcoming Data Barriers in Surgery by Building Infrastructure for Open and Standardised Research

Generating high-quality intraoperative data is one of the most pressing challenges in advancing surgical data science. Real-world operating rooms are bound by strict safety protocols, complex workflows and proprietary systems, which severely limit the ability to capture comprehensive datasets. As a result, the use of data-driven methods in surgery lags behind other disciplines. This persistent lack of access to structured and scalable data has slowed progress across multiple research areas, including 3D reconstruction, behavioural analysis, machine learning and medical imaging.

To address this challenge, the open research data (ORD) project "Surgical Data Science OR-X" was launched in July 2024. It brought together researchers and engineers from the University of Zurich, ZHAW, Zühlke Group and the Digital Medicine Unit of the University Hospital Balgrist. Funded by swissuniversities and the State Secretariat for Education, Research and Innovation (SERI), the project aimed to develop infrastructure and methods for open and standardized surgical data collection. The project was successfully completed in June 2025.

From Concept to Implementation

  • Set-up in the OR-X for the third milestone of the ORD project with the newly developed data hub and the KUKA robot in the background.
  • Vincent Schorp (Data Engineer ORD) explains the first data collected displayed on a Brainlab screen.
  • Using an ultrasonic probe on a KUKA robot, data for the Data Hub is collected on a specimen.
  • An engineer and surgeon operate the KUKA robot, which records ultrasound images.
From Concept to Implementation

The project set out to enable the generation and sharing of large-scale, multimodal surgical datasets, aligned with the principles of Open Science. This included the development of a hardware and software solution for synchronised real-time data capture across diverse surgical technologies, as well as a framework for structuring and standardising the collected data. In addition, a public cloud platform was created to provide access to curated datasets, following the FAIR principles: findable, accessible, interoperable and reusable.

To achieve this, the project relied on the research infrastructure of OR-X, which enabled the integration of technologies such as surgical robotics, augmented reality and artificial intelligence under realistic, but non-clinical conditions. OR-X combined a fully equipped surgical environment with the flexibility to connect experimental devices, synchronise data flows and conduct controlled studies with high reproducibility. This setting allowed the team to develop and test technical solutions that would not have been feasible within the constraints of actual operating rooms.

Beyond the technical setup, the implementation phase also involved iterative validation cycles with external experts from clinical and academic fields. These consultations ensured that the systems developed were not only technologically sound but also relevant and practical for end users in surgical research. By involving stakeholders throughout, the project was able to align its outputs more closely with actual research needs.

Outcomes and Broader Impact

Outcomes and Broader Impact

The project delivered two main results: a Data Collection Framework for automated, standardized data acquisition, and a Surgical Data Cloud Platform that makes selected datasets openly available to the research community. Together, these outcomes help establish new benchmarks for how surgical data can be collected, managed and shared.

Beyond its immediate technical achievements, the project contributes to the creation of discipline-specific standards and methods for surgical data science. It also offers a model for other areas of medicine where data access is similarly constrained. The modular architecture developed through the project, particularly its Data Hubs and Middleware, provides a transferable solution that could support future research infrastructures or even inform clinical data strategies.

In the context of national data initiatives, it demonstrates how Open Research Data initiatives can move beyond theory to produce tangible, reusable resources for science and innovation.