Synthema
Public Deliverables
D1.2 Data and metadata model – Public deliverable
This document provides the Data model and transformation plan to assess the minimal set of variables for each use case and map the two use case model datasets with the Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM)1, the open community data standard selected by the consortium to standardise the structure and content of observational data based on the OHDSI standardised vocabularies, which allow organisation and standardisation of medical terms across the various clinical domains of the OMOP CMD. The document is articulated into 4 sections. After the Introduction, the Data assets section provides a high-level summary of the available data assets within the consortium, including clinical repositories, data sources, types and variables. In the Data modelling section, the data mapping work is reported, starting from an overview of the available data models and standards, a description of the selected OMOP CDM, and the processing of data modelling in the project, including the resulting CDM. In Conclusions, we summarise the results of this process and illustrate the next steps of the work, to be finalised in M30.
D1.3 Data collection and processing report – Public deliverable
SYNTHEMA aims to create a cross-border data infrastructure that supports clinical data standardisation, pseudonymisation, and synthetic and anonymised data generation for haematological diseases (sickle cell disease, acute myeloid leukaemia), in compliance with GDPR. This deliverable describes the data collection and processing activities carried out for the two use cases, including ethical approvals, collection of multimodal data types (clinical, genomic, imaging, metabolomic), and the implementation of the OMOP Common Data Model. Additionally, this report analyses user requirements, data quality aspects, and alignment with RADeep and ENROL registries to ensure harmonised and standardised datasets for future platform development and federated AI validation.
D5.4 Data management plan updated version – Public deliverable
This deliverable provides an update of the Data management plan (DMP) submitted in M6 as D5.1. It incorporates the details provided in D5.1 for the initial plan and captures developments in processes and plans that have occurred in the last 18 months of the project. The deliverable is based on the European Commission template for Horizon 2020 projects1 with some additional answers around future regulatory compliance and sustainability aspects. SYNTHEMA has populated this DMP in line with recommended EC guidelines. This update will be further enhanced towards the end of the project where the plans outlined here will be in a more finalised state and close to full implementation. SYNTHEMA remains novel in that the data assets to be produced will be synthetic in nature. There may yet be a risk of re-identification given that the synthetic data will be generated based on inference from real data. The plan therefore considers this paradigm carefully in its discussion and continues to work alongside other WP5 risk assessment and management approaches. The updated DMP continues to describe the motivations for conducting a DMP, the approach taken and the developments thus far. As per D5.1, D5.4 provides the M24 Update.
D4.1 Synthetic validation framework – Public deliverable
The SYNTHEMA Synthetic Validation Framework (SVF) establishes a rigorous approach for validating synthetic data in clinical research, particularly for diseases like acute myeloid leukaemia (AML) and sickle cell disease (SCD). Designed to ensure data utility while protecting privacy, the SVF supports the safe use of synthetic data in AI healthcare applications. The SVF evaluates synthetic data through three primary dimensions: statistical fidelity, clinical utility, and privacy. Statistical fidelity ensures synthetic data mirrors real data distributions and correlations. Clinical utility assesses the data relevance for insights like survival analysis and mutation frequencies, while privacy metrics evaluate the risk of re-identifying individuals. Data generation approaches were tested, including generative adversarial networks (GANs) and variational autoencoders (VAEs). VAEs showed strong performance in replicating cell characteristics, though rare features require further development. The SVF is essential to SYNTHEMA goal of validating synthetic data for real-world healthcare, ensuring it is accurate, clinically valuable, and privacy compliant. This framework provides a critical foundation for advancing synthetic data use in secure, effective AI-driven medical research.
D7.2 Ethic Design requirements – Public deliverable
This deliverable presents the requirements for an ethical and trustworthy design of SYNTHEMA technologies. This deliverable outlines the regulatory challenges for AI systems in healthcare, discussing the role of regulations as both a barrier and a driver of technology innovation. Then, the seven principles for Trustworthy AI are discussed: these introduce the forthcoming AI Act. The deliverable analyses the forthcoming regulation and the risk categories and gives an overview of the standardisation activities related to the AI Act in the EU. The Value-Sensitive Design methodology is presented, and its implementation in the first year of SYNTHEMA is discussed. In the final section, the discussion on the ethical framework is narrowed down to seven key ethics requirements for the development of the SYNTHEMA technologies.
D7.1 Quality Assurance Guidelines – Public deliverable
The present deliverable constitutes a reference document at consortium partners’ disposal for ensuring the highest quality in the execution of the project, by providing a framework of procedures, guidelines, standards and rules to guarantee the quality of project outcomes (e.g., deliverables, periodic reports, software, infrastructure).
D6.4 Project website – Public deliverable
The SYNTHEMA website (www.synthema.eu) includes key information about the project and aims to communicate to both professionals and wider audiences.
D6.1 Impact Master Plan
This document outlines the project dissemination, communication exploitation strategies for the SYNTHEMA Horizon Europe project.
D5.1 Data management plan – Public deliverable
This deliverable provides the Data management plan (DMP) initial draft for SYNTHEMA. SYNTHEMA has populated this DMP in line with recommended EC guidelines. It will be updated as the project proceeds. SYNTHEMA is novel in that the data assets to beproduced will be synthetic in nature. There may yet be a risk of re-identification given that the synthetic data will be generated based on inference from real data. The plan therefore considers this paradigm carefully in its plans.