Content area
The adoption of the European Health Data Space (EHDS) regulation has made integrating health data critical for both primary and secondary applications. Primary use cases include patient diagnosis, prognosis, and treatment, while secondary applications support research, innovation, and regulatory decision-making. Additionally, leveraging large datasets improves training quality for artificial intelligence (AI) models, particularly in cancer prevention, prediction, and treatment personalization. The European Union (EU) has recently funded multiple projects under Europe’s Beating Cancer Plan. However, these projects face challenges related to fragmentation and the lack of standardization in metadata, data storage, access, and processing. This paper examines interoperability standards used in six EU-funded cancer-related projects: IDERHA (Integration of Heterogeneous Data and Evidence Towards Regulatory and Health Technology Assessments Acceptance), EUCAIM (European Cancer Imaging Initiative), ASCAPE (Artificial Intelligence Supporting Cancer Patients Across Europe), iHelp, BigPicture, and the HealthData@EU pilot. These initiatives aim to enhance the analysis of heterogeneous health data while aligning with EHDS implementation, specifically for the EHDS for the secondary use of data (EHDS2). Between October 2023 and July 2024, we organized meetings and workshops among these projects to assess how they adopt health standards and apply Internet of Things (IoT) semantic interoperability. The discussions focused on interoperability standards for health data, knowledge graphs, the data quality framework, patient-generated health data, AI reasoning, federated approaches, security, and privacy. Based on our findings, we developed a template for designing the EHDS2 interoperability framework in alignment with the new European Interoperability Framework (EIF) and EHDS governance standards. This template maps EHDS2-recommended standards to the EIF model and principles, linking the proposed EHDS2 data quality framework to relevant International Organization for Standardization (ISO) standards. Using this template, we analyzed and compared how the recommended EHDS2 standards were implemented across the studied projects. During workshops, project teams shared insights on overcoming interoperability challenges and their innovative approaches to bridging gaps in standardization. With support from HSbooster.eu, we facilitated collaboration among these projects to exchange knowledge on standards, legal implementation, project sustainability, and harmonization with EHDS2. The findings from this work, including the created template and lessons learned, will be compiled into an interactive toolkit for the EHDS2 interoperability framework. This toolkit will help existing and future projects align with EHDS2 technical and legal requirements, serving as a foundation for a common EHDS2 interoperability framework. Additionally, standardization efforts include participation in the development of ISO/IEC 21823-3:2021—Semantic Interoperability for IoT Systems. Since no ISO standard currently exists for digital pathology and AI-based image analysis for medical diagnostics, the BigPicture project is contributing to ISO/PWI 24051-2, which focuses on digital pathology and AI-based, whole-slide image analysis. Integrating these efforts with ongoing ISO initiatives can enhance global standardization and facilitate widespread adoption across health care systems.
Full text
Introduction
Cancer is one of the main causes of death in Europe and worldwide, after cardiovascular diseases. According to the World Health Organization (WHO), cancer is the second leading cause of death and morbidity in Europe, with more than 3.7 million new cases and 1.9 million deaths each year [1]. In response to the urgent need to renew European political commitment to tackle cancer, Europe’s Beating Cancer Plan (EBCP; 2021-2027) was launched and structured around 4 key action areas: prevention, early detection, diagnosis and treatment, and the quality of life of patients with cancer and cancer survivors [2]. Since 2021, the European Commission (EC) has supported collaborative projects focused on cancer diagnostics and treatment using high-performance computing and artificial intelligence (AI). To maximize the potential of data and digitalization, the EBCP also addressed the interactions and alignment of cancer data projects and initiatives with the European Health Data Space (EHDS). In autumn 2023, the EC organized 3 online workshops on the reuse of health data resources in the field of cancer research and recently published the results of these workshops [3], where “fragmentation and the lack of standardization in metadata, data storage, access, and processing” were identified as key challenges facing data-driven cancer projects. The nascent infrastructure for the application of AI in medical imaging (European Cancer Imaging Initiative [EUCAIM]) reported on the experiences of 5 projects developing big data infrastructures that will enable European, ethical, General Data Protection Regulation (GDPR)–compliant, quality-controlled, and cancer-related medical imaging platforms, where both large-scale data and AI algorithms will coexist [4]. These projects include the following:
Table 1 summarizes the list of projects that participated in the EC workshop entitled “Landscaping data-driven projects and initiatives in the cancer field–rationale and directions for better collaboration and integration” [3], as well as the established project-level synergies and collaborations among ongoing projects.
The workshop also addressed the current challenges facing data-driven cancer projects [3], gaps existing in existing standards, and recommendations for future semantic interoperability (given in Table 2).
Table 1.Collaborations and synergies among the European Union’s existing cancer projects.
| Collaboration scope | Synergy projects |
| Data representation and interoperability | HealthData@EU pilot and CanSERV |
| Infrastructure and services for benchmarking | EOSC4Cancer, EUCAIMa, and TEF-Health |
| Federated data infrastructure and data structure | EUCAIM, EOSC4Cancer, and GDI |
| Secure processing environment | SOLACE, EUCAIM, CanScreen-ECIS, IDERHAb, and Optima |
aEUCAIM: European Cancer Imaging Initiative.
bIDERHA: Integration of Heterogeneous Data and Evidence Towards Regulatory and Health Technology Assessments Acceptance.
Table 2.Main challenges facing European Union cancer projects and recommendations.
| Challenges [3] | Gaps in existing standards | Recommendations |
| Fragmentation and lack of standardization in metadata, data storage, access, and processing. | The DCAT-APa health extension is still under development by the HealthData@EU pilot. | TEHDASb JAc recommendations to enhance interoperability within HealthData@EU—a framework for semantic, technical, and organizational interoperability. |
| Lack of access to diverse and high-quality datasets. | Data quality framework for primary care data sources is not sufficiently addressed in the EHDSd framework. | Data quality frameworks provided by TEHDAS, EMAe, and QUANTUM projects. |
| Evolving legal landscape | Relevant standards to the AIf Act are under development. | Participating in developing or extending the relevant standards. |
aDCAT-AP: DCAT Application Profile for Data Portals in Europe.
bTEHDAS: Towards the European Health Data Space.
cJA: joint action.
dEHDS: European Health Data Space.
eEMA: European Medicines Agency.
fAI: artificial intelligence.
HSbooster Health Project Cluster in Cancer
In this work, we focus on the interoperability challenges and existing gaps in health care standards building on the challenges and synergies highlighted in the EC workshop [3]. Through the EC European Standardization Booster (HSbooster.eu) initiative, we create synergy among six cancer data-driven projects (ie, IDERHA [Integration of Heterogeneous Data and Evidence towards Regulatory and Health Technology Assessments Acceptance], EUCAIM, ASCAPE [Artificial Intelligence Supporting Cancer Patients Across Europe], iHelp, BigPicture, and the HealthData@EU pilot project) by using health standards. These 6 innovative projects aim to transform digital health in oncology by leveraging advanced technologies and collaborative frameworks for enhancing cancer diagnosis, treatment, and research, improving patient outcomes, and accelerating scientific advancements. We initially created a template for the EHDS for the secondary use of data (EHDS2) interoperability framework based on the new European Interoperability Framework (EIF). The recommended standards and governance model from the joint action (JA) Towards the European Health Data Space (TEHDAS) [5] for the secondary use of data were then used to harmonize the standards in the template (given in Textbox 1).
Starting October 2023, we conducted several meetings and workshops among the 6 projects to elucidate how they adopt health standards and Internet of Things (IoT) semantic interoperability. This included examining interoperability standards for health data, knowledge graphs–related technologies, the Smart Applications Reference Ontology (SAREF), the data quality framework (DQF), patient-generated health data (PGHD), AI reasoning, federated approaches, security, and privacy.
As per the TEHDAS recommendations, we compared the health-standardized models, ontologies, and terminologies used in these projects, including, Health Level 7 (HL7) Fast Healthcare Interoperability Resources (FHIR), Open Health Data Science and Informatics (OHDSI), Observational Medical Outcomes Partnership (OMOP)–Common Data Model (CDM), Digital Imaging and Communications in Medicine (DICOM), International Organization for Standardization (ISO) Technical Committee (TC) 215, and World Wide Web Consortium (W3C) Data Catalog Vocabulary (DCAT).
As EHDS is still evolving and its implementation is still under development, this study aims to examine how the 6 projects implement the recommended EHDS2 standards. Consequently, we introduce a new template to support the EHDS2 interoperability framework and highlight the results of comparing data standards across projects. By summarizing the lessons learned, we provide recommendations for future directions in EHDS2 implementation.
Textbox 1.European Health Data Space for European Interoperability Framework.
European Health Data Space
Figure 1 shows the process of creating the template for the EHDS2 interoperability framework.
Exploring How EHDS Adopts the New EIF Architecture and Principles
Recent studies showed the importance of adopting the EIF in establishing the EHDS interoperability framework [7,10,11] (as shown in Textbox 1). In addition, several studies have investigated the interoperability requirements for heterogeneous health information systems, as well as the associated health care standards [12,13].
Aligning the TEHDAS Results to the New EIF
The TEHDAS JA involves 25 European countries in developing the principles that will shape EHDS2 by providing guidance and recommendations on interoperability, data quality, and standards (as shown in panel 2 in Textbox 1). In 2023, the TEHDAS JA published a report titled “Options for governance models for the European Health Data,” which discusses EHDS governance using the EIF [14]. In addition, they assessed 19 standards that provide a layer of semantic interoperability, supporting the cataloging of data sources and the exchange of data between different nodes [15].
Mapping the TEHDAS Recommended Standards and Governance Model to the New EIF
Following the recommendation of the EC workshops on landscaping data-driven projects and initiatives in cancer [3], we adopted the TEHDAS results on recommended standards for EHDS2 interoperability, as well as the principles of data quality frameworks. We then mapped these results to the new EIF interoperability framework and linked the data quality framework to the corresponding ISO standards. Figure 2 shows how the TEHDAS results are mapped to the new EIF interoperability framework. Notably, TEHDAS framed the recommended EHDS2 interoperability into 3 categories of standardization:
The EIF does not classify standards according to interoperability layers (technical, semantic, legal, and organizational). However, we need to consider the relevant standards for data privacy and quality assessment [10].
Furthermore, the work is a foundation for creating an interactive tool for the EHDS2 Interoperability Framework, which has been submitted to JMIR as part 2 (Towards EHDS2 interoperability framework: An interactive EIF-based standards compliance toolkit for AI-driven projects).
Linking DQF to the Relevant ISO Standards
In May 2024, the Big Data Value Association (BDVA) published a study entitled “Elevating Data Quality A Paradigm Shift for Data Spaces and AI Needs” to explore the relationships between data quality and data spaces with respect to the AI Act [17]. The study proposed that the data quality and utility label for EHDS should comply with data documentation, technical quality, data quality management processes, coverage, access and provision, and data enrichment procedures.
From the EHDS2 perspective, the TEHDAS JA provided a generic DQF [18], which includes both technical quality elements and six utility dimensions: relevance, accuracy and reliability, coherence, coverage, completeness, and timelines. This was followed by the publication of associated recommendations affecting data quality and utility implementation in HealthData@EU [19]. Based on these deliverables, the European Medicine Agency (EMA) published its own DQF for EU medicines regulation [20]. This publication provides an analysis of the data quality actions and metrics, as well as a maturity model, to guide the evolution of automation to support data-driven regulatory decision-making (as shown in Figure 3) [20].
Ensuring the accuracy, completeness, consistency, and reliability of cancer research data is of utmost importance, and adherence to data quality standards plays a crucial role in achieving this goal. These standards not only help maintain data integrity but also promote data interoperability through the use of standardized data models and vocabularies. This enables seamless data exchange and integration across different projects and platforms. In addition, adhering to these standards supports informed decision-making by providing high-quality data for clinical decisions, research insights, and policymaking related to cancer treatment and prevention. Compliance with international legal and regulatory requirements is also ensured, upholding proper data governance and ethical data usage. Finally, establishing a unified framework for data quality encourages collaboration among various stakeholders within the cancer research community. We compare data quality standards in Table S4 in Multimedia Appendix 1.
Landscape of the Involved Projects and Used Standards
In this study, we used the EHDS2 interoperability framework template to compare 6 projects with cancer use cases where AI is applied (as shown in panel 4 in Textbox 2; Table 3). The selected projects vary in scope, cancer domain, categories of health data, scale of infrastructure, AI implementation approach, and time spans [21].
We focused on semantic interoperability standards (as shown in Textbox 3), recommended by the TEHDAS JA and related ISO standards. The analysis explored how health standards support health ontologies, for example, how HL7 FHIR supports the Findability, Accessibility, Interoperability, and Reusability (FAIR) principles through the FHIR for FAIR implementation guide [28].
Textbox 2.Introduction to six key projects advancing oncology through digital health.
We explore 6 innovative projects that are transforming digital health in oncology. By leveraging advanced technologies and collaborative frameworks, these projects aim to enhance cancer diagnosis, treatment, and research, improving patient outcomes, and accelerating scientific advancements:
| Project | Duration | Cancer domain | AIa or MLb | Comments |
| IDERHAc | 2023-2028 | Lung cancer | Federated machine learning | Under implementation |
| BigPicture | 2021-2027 | Pan-cancer whole slide images | Central repository for digital pathology and platform for AI development | Ongoing |
| EUCAIMd | 2023-2026 | Pan-cancer images | Federated Research Infrastructure | Ongoing |
| iHelp | 2021-2024 | Pancreatic cancer | Explainable AI, deep neural networks, predictive algorithms, ML techniques, and federated queries on distributed infrastructures | Ended on June 30, 2024 |
| ASCAPEe | 2020-2023 | Breast and Prostate cancer | Explainable AI, federated deep learning, and ML on homomorphically encrypted data | Finished |
| HealthData@EU pilot | 2022-2024 | Colorectal cancer and other nonrelated cancer use cases | Federated query in noncancer use case: machine learning for analysis of care pathways | Pilot for EHDS2 |
aAI: artificial intelligence.
bML: machine learning.
cIDERHA: Integration of Heterogeneous Data and Evidence Towards Regulatory and Health Technology Assessments Acceptance.
dEUCAIM: European Cancer Imaging Initiative.
eASCAPE: Artificial Intelligence Supporting Cancer Patients Across Europe.
Textbox 3.Advancing semantic interoperability in digital health: a landscape of standards used in the involved projects.
Our attention is drawn to the critical role that standards play in examining semantic interoperability within the digital health realm. These standards form the foundation for seamless data exchange, ensuring consistency and coherence across various health care platforms. The main standards are as follows:
* ISO TC 215 and CEN TC 251 Health Informatics: These cover a range of ISO standards, such as 13606 (5 parts), 27269, and 29585, alongside the Health Informatics Service Architecture (HISA)–compliant data models.
Abbreviations of semantic metadata or terminologies used by standards
To identify synergies among the 6 projects, we used the EHDS2 interoperability framework template to analyze adopted standards, highlighting similarities and differences in the implementation approach (detailed results given in Tables S1-S7 in Multimedia Appendix 1).
Mapping the Projects to the Created Template for the EHDS2 Interoperability Framework
The EHDS2 interoperability framework template identified key standards and technologies for the implementation of EHDS2, including HL7 FHIR; DICOM; OMOP-CDM; upcoming standards developed by ISO TC 215, ISO TC212, CEN TC 251; and ontology technologies, including W3C DCAT-AP. The key standards used in 6 projects, their focus areas, and planned future implementations are summarized in Table 4.
Table 4.Overview of the used, innovated, and planned health data standards in the projects (using the template of the European Health Data Space for the secondary use of data [EHDS2] interoperability framework).
| Project | Used standards | Planned or future standards |
| IDERHA | HL7a FHIRb, DICOMc, and OMOP | ISOd TCe 215, DCAT-APf: HealthDCAT-AP |
| BigPicture | DICOM, SNOMEDg, and ICD | ISO TC 212 (digital pathology and AIh), and Kidney Biopsy Codes |
| EUCAIMi | Using its own hyperontology and common data model based on: DICOM, DICOM Seg, OMOP, FHIR, mCODE, DCAT-AP-Health, OSIRIS Ontologies used so far: LOINCj, SNOMED, UCUMk, RADLEXl, ICDO3, ICD-10m, CPT4, ICD10PCS, ATCn, NCITo, Birnlexp, NAACRq, Cancer Modifier | —r |
| iHelp | HL7 FHIR, OMOP-CDM, ISO 27799:2016, SNOMED, LOINC, ICD-9, ICD-10, UMLS, SPARQL, RDFSs, RxNorm | — |
| ASCAPE | HL7 FHIR, ISO/CEN 13606, LOINC, SNOMED | — |
| HealthData@EU pilot | Observing and collecting standardization efforts rather than direct implementation | The first version of the Health DCAT-AP metadata standard is being developed in the project |
aHL7: Health Level 7.
bFHIR: Fast Health Interoperability Resources.
cDICOM: Digital Imaging and Communications in Medicine.
dISO: International Organization for Standardization.
eTC: Technical Committee.
fDCAT-AP: DCAT Application Profile for Data Portals in Europe.
gSNOMED: Systematized Medical Nomenclature for Medicine.
hAI: artificial intelligence.
iEUCAIM: European Cancer Imaging Initiative.
jLOINC: Logical Observation Identifiers Names and Codes.
kUCUM: Unified Code for Units of Measure.
lRADLEX: Radiology Lexicon.
mICD-10: International Classification of Diseases, Tenth Revision, Clinical Modification.
nATC: Anatomical Therapeutic Chemical classification system.
oNCIT: National Cancer Institute Thesaurus.
pBirnlex: Biomedical Research Integrated Domain Group.
qNAACR: North American Association of Central Cancer Registries.
rNot applicable.
sRDFS: Resource Description Framework Schema.
Main Findings and Lessons Learned
All projects support the recommended standards identified by the TEHDAS JA for EHDS2 interoperability. Although the DICOM standard is widely used for collecting, storing, and transferring medical imaging data, it lacks important information required to identify relevant images because DICOM metadata are not standardized [37]. To overcome this challenge, the EUCAIM project CDM is built upon the FHIR resources ImagingStudy and ImagingSeries, the Medical Imaging–CDM extension of the OMOP-CDM, the ProCAncer-I imaging extension [38], and the OSIRIS imaging component. In this way, proper integration of imaging and clinical data was provided in alignment with the Integrating the Healthcare Enterprise [39].
The BigPicture project selected the DICOM format as a standard for archiving whole-slide images (WSI) to support back-and-forth conversion of proprietary file formats. The BigPicture profile for DICOM WSI is based on work by the DICOM pathology working group WG26. The format is designed to allow efficient storage of a large number of annotations, for example, generated by AI algorithms [40]. The project developed a Python tool (Wsidicom) to serve as a reference implementation of a DICOM WSI reader [41]. In addition, two tools, Opentile and Wsidicomizer, were developed to read other WSI formats and convert them to DICOM resulting in faster format conversion and maintained image quality [42].
To ensure structured data exchange and harmonized-consistent metadata interaction for data not captured in DICOM, BigPicture has developed additional metadata standards that define and comprise a set of substandards: (1) the data model (Common Mandatory Metadata Structure[CMMS]) comprising of all metadata, (2) a set of mandatory information that must be provided in relation to CMMS entities and their relations to each other, (3) a metadata file format that is used within BigPicture (flexible metadata file exchange format), and (4) a standard file structure of datasets containing metadata and data files that can be found on the repository. Standards are based on the European Genome-phenome Archive (EGA), and where possible existing standards have been incorporated (SNOMED, ICD for clinical data, Standardization for Exchange of Nonclinical Data [SEND] terminology, and International Harmonization of Nomenclature and Diagnostic Criteria (INHAND) nomenclature for nonclinical data). Results from BigPicture, such as metadata formats or quality control criteria, will feed into the ISO/AWI 24051-2 “Medical laboratories—Part 2: Digital pathology and artificial intelligence (AI)-based image analysis.”
Regarding the metadata standards, the HealthData@EU pilot published a landscape analysis of available metadata catalogs and metadata standards [43]. The pilot also develops an extension of DCAT-AP: HealthDCAT-AP that will be adopted by EHDS2 projects, like IDERHA [44]. This standard could be adopted later as the mandatory metadata standard foreseen under the EHDS regulation. The EUCAIM project designed its Hyper Ontology [45] using FHIR, UMLS, SNOMED-CT, and OMOP-CDM vocabularies.
While data from wearables are available in high volumes, a substantial amount of data is lost because the usability of the data is governed and limited by proprietary data formats from an increasing number of manufacturers, which shows the necessity of standardization to enable interoperability [46]. Among the six projects and initiatives, IDERHA, iHelp, and ASCAPE use sensors and wearables technologies. The iHelp project uses the holistic health records (HHR) model to enable the aggregation of data from different sources, sensors, and online platforms to support the seamless integration of multiple health dimensions [47].
The HL7 FHIR standard-compatible data structures were used in the context of the iHelp project to integrate primary and secondary data and to compile them into the holistic health records FHIR model [47]. FHIR can be used for clinical data and also for streaming data from sensors [48]. The ASCAPE project successfully integrated EN/ISO 13606–standardized extracts from a patient mobile app into an electronic health record [49]. IDERHA plans to extend the OMOP-CDM to address the PGHD, including patient-reported outcomes and patient-reported experience measures.
The BigPicture project highlighted that a dedicated standardization project focused on digital pathology is currently missing. It addressed the need for ISO/AWI 24051-2 “Medical laboratories—Part 2: Digital pathology and artificial intelligence (AI)-based image analysis” (under development, April 2024) that will build on ISO 20166-4:2021 “Molecular in vitro diagnostic examinations Specifications for pre-examination processes for formalin-fixed and paraffin-embedded (FFPE) tissue Part 4: In situ detection techniques” [50]. Figure 4 summarizes the main findings of the projects and how they can feed each other.
It is interesting to notice that the ETSI SmartM2M SAREF extensions for health and aging well (SAREF4EHAW) and wearables (SAREF4WEAR) are not known and not used on those six projects that we compared, even projects using IoT technologies with sensors, devices.
Key reasons for this gap include insufficient promotion, perceived relevance, integration challenges, and the presence of competitive standards. ETSI should enhance outreach, showcase successful implementations, collaborate with key projects, and simplify integration processes to improve adoption.
The other aspects, AI and federated learning, security and privacy, and data quality are covered by several approaches and standards. The adoption of federated standards and AI in cancer research projects across Europe represents a shift toward collaborative yet privacy-preserving medical research. These initiatives are essential for creating robust, scalable, and interpretable AI models that can significantly advance early detection, treatment, and overall patient care in oncology. The projects leverage various AI standards to achieve these goals. Some of the key AI standards being used have been included in Table S7 in Multimedia Appendix 1. The Regulation (EU) 2018/1725, also known as GDPR, covers the fundamental requirements for access, sharing, and storage of personal data, including health data. The usage of raw health data with personal details included for research purposes remains controversial in terms of consent and audit logging. Furthermore, there is a need for standardized security and data-sharing models. Finally, conformance guidelines to ensure compliance with regulations (eg, AI Act) for manufacturers developing medical products are still needed.
Creating Synergy Among the Involved Projects
We remarked on similarities and differences as we navigated the implementation of various standards among the projects. Despite the projects’ varied focuses, each project prioritizes interoperability, relying on various standards. This collective effort promotes knowledge exchange and innovation and fosters a digitally unified health care ecosystem for EHDS2. We also observed a trend towards common standards like HL7 FHIR and DICOM. While projects may vary in domain-specific standards and implementation nuances, leveraging established standards offers numerous advantages. This includes improved interoperability, streamlined development processes, scalability, and knowledge sharing.
Accordingly, intensive discussions accelerated the creation of the synergy among the groups that will be introduced in the Medical Informatics Europe 2024 conference [21]. In addition, both the HealthData@EU pilot and the IDERHA project were addressed as EHDS use cases in the white paper on “IoT/Edge Computing and Health Data and Data Spaces” published by the Alliance for IoT and Edge Computing Innovation [51], and both projects also shared expertise on informing EHDS2 development. Experience from completed projects working with standards helps ongoing projects to decide on which standards to focus on, learn from their limitations, etc. For example, BigPicture designed a new standard ISO/AWI 24051-2.
Conclusion and Outlook
Standardization and legalization are the main pillars of EHDS. The HSbooster.eu initiative enabled intensive analysis of these aspects among six data-driven EU projects focusing on cancer. We aimed to create synergy among these projects to share the lessons learned in standards and legal implementation, sustainability of these projects, and harmonization with these with the EHDS2 implementation. Furthermore, we are involved in the development of standards such as ISO/IEC 21823-3:2021 “IoT - Interoperability for IoT Systems - Part 3 Semantic interoperability” (as editors), and since there is no ISO Standard for digital pathology and AI-based analysis of (whole slide) images for medical diagnosis, BigPicture is involved in ISO/PWI 24051-2 “Medical laboratories—Part 2: Digital pathology and artificial intelligence (AI)-based image analysis.”
Future work requires integration of results from TEHDAS 1 & 2 and HealthData@EU pilot as well as other EHDS supporting projects, such as QUANTUM [52], which is seeking to develop the EHDS data quality and utility label. Additionally, enhancements to the EHDS2 interoperability framework template are needed, with more standards covering health ontologies, wearables, personal devices, data quality, as well as the usage of FAIR principles. The template can help researchers choose appropriate standards for their projects, thereby reducing time and effort in standard selection and implementation and improving EHDS2 implementation. Moreover, the template will be used in designing and creating an interactive toolkit for the EHDS2 interoperability framework. In this way, the existing and future projects can ensure alignment with governance and interoperability requirements for EHDS2.
EHDS has recently gained additional relevance since the European AI Act that has been approved by the European Parliament by resolution on March 13, 2024, explicitly refers to the EHDS as needing to “facilitate non-discriminatory access to health data and the training of AI algorithms on those data sets, in a privacy-preserving, secure, timely, transparent and trustworthy manner, and with an appropriate institutional governance.”
Acknowledgments
AG declares funding from HSBOOSTER 101058391, StandICT.eu 2026 101091933. The IDERHA (Integration of Heterogeneous Data and Evidence Towards Regulatory and Health Technology Assessments Acceptance) project is supported by the Innovative Health Initiative (IHI) Joint Undertaking (JU) under grant agreement 101112135. The JU receives support from the European Union’s Horizon Europe research and innovation program and life science industries represented by COCIR, EFPIA/Vaccines Europe, EuropaBio, and MedTech Europe. The BigPicture project has received funding from the Innovative Medicines Initiative 2 JU under grant agreement 945358. This JU receives support from the European Union’s Horizon 2020 research and innovation program and EFPIA. The iHelp project received funding from the European Union's Horizon 2020 research and innovation program under grant agreement 101017441. The ASCAPE (Artificial Intelligence Supporting Cancer Patients Across Europe) project received funding from the European Union’s Horizon 2020 research and innovation program under grant agreement No 875351. The EUCAIM (European Cancer Imaging Initiative) project is cofunded by the European Union under grant agreement 101100633. The HealthData@EU pilot is cofunded by the European Union.
Authors' Contributions
AG and RH contributed to conceptualization, methodology, and writing – original draft. SA and PG assisted with methodology and writing – original draft. GM, RvN, KZ, IEN, GD, SN, LM-B, PM, SD, SA, MJ, IA, EGA, PH, IB, and AB contributed to validation and writing – review & editing.
Conflicts of Interest
None declared.
------------
Abbreviations
| AI: artificial intelligence |
| AI4HI: Artificial Intelligence for Health Imaging |
| ASCAPE: Artificial Intelligence Supporting Cancer Patients Across Europe |
| BDVA: Big Data Value Association |
| CDM: common data model |
| CMMS: Common Mandatory Metadata Structure |
| DCAT: Data Catalog Vocabulary |
| DCAT-AP: DCAT Application Profile for Data Portals in Europe |
| DICOM: Digital Imaging and Communications in Medicine |
| DQF: data quality framework |
| EBCP: Europe Beating Cancer Plan |
| EC: European Commission |
| EHDS: European Health Data Space |
| EHDS2: European Health Data Space for the secondary use of data |
| EIF: European Interoperability Framework |
| EMA: European Medicine Agency |
| EGA: European Genome-Phenome Archive |
| EU: European Union |
| EUCAIM: European Cancer Imaging Initiative |
| FAIR: Findability, Accessibility, Interoperability, and Reusability |
| FHIR: Fast Healthcare Interoperability Resource |
| GDPR: General Data Protection Regulation |
| HL7: Health Level 7 |
| IDERHA: Integration of Heterogeneous Data and Evidence Towards Regulatory and Health Technology Assessments Acceptance |
| IEC: International Electrotechnical Commission |
| INHAND: International Harmonization of Nomenclature and Diagnostic Criteria |
| IoT: Internet of Things |
| ISO: International Organization for Standardization |
| JA: joint action |
| OHDSI: Observational Health Data Sciences and Informatics |
| OMOP: Observational Medical Outcomes Partnership |
| PGHD: patient-generated health data |
| SAREF: Smart Applications Reference Ontology |
| SEND: Standardization for Exchange of Nonclinical Data |
| SNOMED CT: Systematized Medical Nomenclature for Medicine–Clinical Terminology |
| TC: Technical Committee |
| TEHDAS: Towards the European Health Data Space |
| WHO: World Health Organization |
| WSI: whole-slide images |
| W3C: World Wide Web Consortium |
------------
Edited by A Coristine; submitted 09.09.24; peer-reviewed by A Billis, Z Hou, FA Causio; comments to author 18.11.24; revised version received 10.12.24; accepted 31.01.25; published 24.03.25.
© 2025. This work is licensed under https://creativecommons.org/licenses/by/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.