Research in Orthopedics

Connecting Orthopedic Medicine and Data-Driven Research

The Promise of Digital Medicine and AI: The ability to create digital twins from multimodal data. For this, data integration in a sustainable and transparent framework is required.

The Digital Medicine Unit is developing a sustainable digital framework that pursues two main objectives: firstly, to increase the efficiency of research projects by supporting data management, and secondly, to develop and validate trustworthy artificial intelligence models and digital tools to support clinical processes.

With the increasing digitalization of medicine, patient data have become omnipresent, with access to everlarger sets of clinical information being provided. Powerful computational approaches allow us to capitalize on this data and to generate novel medical insights by using predictive models for disease outcome and treatment efficacy. There is little doubt that we are seeing the emergence of data-driven medicine, in which data and artificial intelligence (AI) shape how we diagnose and treat patients. As a result, patients should greatly benefit from more personalized health care approaches and physicians will receive operational and decison-making support.

In practice, however, getting access to and using clinical data for research and innovation is a complex endeavor that traditionally involves many challenges. To overcome these issues, a uniform solution is required, rather than individual systems creating multiple silos. This means that the clinical database of tomorrow needs to be centralized, harmonized, and searchable by every party involved. Moreover, it must be capable of exchanging data efficiently while still adhering to international standards. So far, an innovative and agile environment that enables fast iterations in the R&D process of medical digital applications does not yet exist.

In this regard, the Balgrist University Hospital (UKB) has positioned itself at the forefront of innovation. It has initiated a proof of concept to demonstrate how an interface between a trusted cloud research environment with the clinical IT framework could be the right foundation for a modern clinical research environment. In close collaboration with multiple technological partners such as Microsoft, Cisco, and NVIDIA, the hybrid nature of this novel is aimed at leveraging cloud services on the one hand, and on the other to promote its eff ective integration into conventional clinical systems. This in turn addresses the clinical translation of novel digital technologies — in particular of AI-based tools. In fact, although in recent years the wave of breakthroughs in AI has seen applications being developed rapidly demonstrating remarkable performance in various aspects, the clinical adoption of these technologies brings with it additional challenges.

Objectives

  • Develop a cloud-based trusted research environment supporting Balgrist University Hospital in overcoming the regulatory hurdles to leverage cloud services and offering researchers an experimental basis to develop, test, and validate datadriven analysis tools.
  • Create a cloud-specific platform addressing data security, data privacy, pseudo- and anonymization, attribute-based privacy measures, and ethical norms.
  • Deploy machine learning (ML)-assisted applications in a collaborative environment to accelerate the creation and testing of multimodal digital twins (DTs) use cases, e.g. digital mirroring of anatomical structures for 3D surgical planning.
  • Identify relevant clinical needs with UKB physicians and translate valuable data-driven applications back to the clinic. An example is the standardization of medical reports using natural language processing (NLP) or novel large language model (LLM) methods.
Anatomical 3D Models
Anatomical 3D Models of Multiple Structures: Thanks to the digital framework, we can automatically process medical images with multiple state-of-the-art AI models. In this image: example of segmentation to create anatomical 3D models of multiple structures.

In conclusion, the main goal of the DMU Team is to offer a local, cloud-based, and collaborative platform accelerating the development of clinical AI tools and DTs. Furthermore, the DMU aims to address topics of trustworthiness and explainability using both a systemic and interdisciplinary approach. The envisioned digital ecosystem will enable eff ective interfaces with multiple projects in which the DMU is actively involved:

  • The LOOP BioMedical Informatics Platform - BMIP
  • The LOOP BMI Imaging Platform
  • The Digital Society Initiative (DSI) Infrastructure & Labs Health Data Repository
  • The Digital Health Zurich project by the Digitalization Initiative of the Zurich Higher Education Institutions (DIZH)
  • The Imaging Hub project by the Personalized Health and Related Technologies (PHRT) initiative at ETH initiative
  • The Health Data Repository and the OR-X projects at UKB

Digitale Plattform für datengesteuerte Forschung und klinische Anwendung

Mit der zunehmenden Digitalisierung in der Medizin sind Patientendaten in der Form immer grösserer Datensätze klinischer Informationen weitgehend zugänglich geworden. Leistungsstarke rechnergestützte Ansätze ermöglichen es uns, diese Daten zu nutzen und neue medizinische Erkenntnisse zu generieren, indem wir präzisere Vorhersagemodelle für Krankheitsverläufe und die Behandlungseffizienz verwenden. Das neue datengesteuerte Paradigma im Gesundheitswesen schafft Möglichkeiten für Patientinnen und Patienten sowie ärztliches Fachpersonal, von stärker personalisierten Gesundheitsansätzen zu profitieren, die auf digitalen Tools basieren. In der Praxis sind jedoch der Zugang zu und die Nutzung von klinischen Daten für die Forschung und die Innovation eine nicht alltägliche Angelegenheit, was traditionell viele Herausforderungen mit sich bringt. Die Universitätsklinik Balgrist konnte dank der Unterstützung mehrerer technologischer Partner einen Proof of Concept initiieren, um zu demonstrieren, wie eine Schnittstelle mit einer vertrauenswürdigen digitalen Umgebung das Fundament für eine moderne klinische IT-Umgebung bildet. Hauptziel dieser Digital Medicine Unit (DMU) ist es, eine kollaborative Plattform anzubieten, die die Entwicklung klinischer KI-Tools beschleunigt und Themen der Vertrauenswürdigkeit und der Erklärbarkeit mit einem systemischen und einem interdisziplinären Ansatz angeht. Das digitale Ökosystem hat das Potenzial, die digitale Kommunikation zwischen klinischen Systemen und Forschungsprojekten zu vereinfachen und effektive Schnittstellen mit mehreren Projekten herzustellen, an denen die DMU aktiv beteiligt ist.

At a Glance

Health Data Repository

Key Collaborators
Project lead: Dr. Sebastiano Caprara
Benedikt Herzog
Nicholas Bünger
Dr. Rui Santos
Dr. Michael Rebsamen

Departments and Partners
Balgrist University Hospital: Digital Medicine Unit (DMU), ROCS, OR-X, Dept. of Radiology
Balgrist Foundation
University of Zurich, Digital Society Initiative
ZHAW, Center for Computational Health
The LOOP Zurich, BioMedical Informatics Platform (BMIP)

Clinical Relevance
The new platform integrating novel digital tools will support clinical processes and will enable data-driven clinical research.

Further Information
> Balgrist Research and Innovation Node
> Digital Society Initiative University of Zurich
> The Loop Zurich
> Digital Health Zurich