Genome-Centric Multimodal Data Integration in Personalised Cardiovascular Medicine.

Goethe University Frankfurt

The Centre for Sudden Cardiac Death and Familial Arrhythmias at the University Hospital Frankfurt consists of an interdisciplinary team of geneticists, cardiologists and pathologists. Together we undertake examinations into the possible underlying causes of sudden cardiac death in the young (SCDY) and we offer interdisciplinary support for the family of the deceased. We are particularly interested in sudden cardiac death in the young (SCDY) and the rare genetic disorders that can lead to (lethal) cardiac arrhythmias. Our research interests lie in the improvement and targeted diagnostic possibilities, risk stratification and therapy in susceptible individuals, working towards the prevention of SCDY in affected families. Moreover, our registry RESCUED (REgistry for Sudden Cardiac and UnExpected Death) offers a unique set of data, since it has a holistic, pedigree-based approach to SCDY.

The Institute of Medical Informatics (IMI) connects clinicians, researchers, and computer scientists at the University Hospital Frankfurt to develop innovative IT solutions for healthcare and research. Through several (inter-)national projects, IMI has extensive experience in managing and integrating clinical data, including standardization and harmonization of data for federated access. IMI played an integral role in establishing the local data integration centre which aims to make routine healthcare data available for research locally and across sites. With the open-source software framework OSSE and the underlying metadata repository, IMI is developing a flexible solution for creating interoperable patient registries, including the RESCUED registry.

Our use case “Integrating registry data into risk assessment algorithms for prevention of sudden cardiac death in the young” is based on data collected in the RESCUED registry (Registry for Sudden Cardiac/Unexpected Death), comprising clinical data, autopsy data and sequencing data from >100 deceased and their families. To comply with the standards set forth in NextGen and to enable federated machine learning approaches, we will ensure compatibility of our data and associated metadata with these standards and take the necessary steps for data curation. Mapping to standard data models or concepts such as CDISC or OMOP will be backed by the underlying registry software OSSE (Open-Source Registry System for Rare Diseases) and its central metadata repository, containing structured and machine-readable data element specifications. In addition, we will explore what additional data from the deceased and the affected families might be relevant for risk assessments, such as ECGs or extended genetic analyses. In combination with the available registry data, this can be used to demonstrate the effective integration of multimodal data into the developed risk assessment algorithms. Furthermore, we will work closely with the Data Integration Center at the University Hospital Frankfurt in in setting up and integrating tools for federated data access provided by NextGen.