Release 2026-5
Read in GermanDutch
Read in DutchDutch
Overview
Classification: Major Feature Release
This release moves the Founda platform onto Apache Tomcat 11 and Java 21 and adds support for Windows Server 2025 and SQL Server 2025, keeping on-premise and cloud installations on a supported, actively patched foundation.
Image Prefetch gains more precise rule matching and a more actionable monitoring dashboard, so prefetch delivers the studies that matter and failures can be resolved directly by operational teams.
The Founda Health Portal now supports managing additional patient identifiers, the Integration Engine can read and reuse values from JSON responses in flows, and imaging records shared through MedMij are aligned with the Nictiz profile for the Dutch market. Reported defects have been resolved across the platform, and ongoing maintenance and security work continues across all deployments.
Upgrade notice: This release runs on Apache Tomcat 11 and requires Java 21. Refer to the TRM for detailed steps.
What's New – All Markets
Platform upgrade: Apache Tomcat 11 and Java 21
The Founda application stack has moved from Apache Tomcat 9 to Tomcat 11 and from Java 17 to Java 21, well ahead of the Tomcat 9 end of life in March 2027.
- Current, supported runtime: all components now run on Jakarta EE 11 (Servlet 6.1, JSP 4.0) on Java 21.
- No functional change: the upgrade was validated through integration and IHE compliance testing, backwards-compatibility testing and smoke tests.
What this means for you
Your installation runs on a supported Tomcat line that continues to receive security fixes, with no change to how the platform behaves for users. Plan the move to Java 21 using the steps in the TRM.
Windows Server 2025 support
Founda is now validated on Windows Server 2025, so hospital infrastructure teams can plan their server refresh around Microsoft's lifecycle dates.
- Supported combinations: Windows Server 2025 with SQL Server 2022, and Windows Server 2025 with SQL Server 2025.
- Validated end-to-end: fresh installations were run through the full automated release regression suite, TLS and certificate handling.
What this means for you
Hospitals can plan a move to Windows Server 2025 with either SQL Server 2022 or SQL Server 2025. Validation covered new installations. In-place upgrades of the operating system or SQL Server follow Microsoft's upgrade guidance and were not part of this validation.
Image Prefetch enhancements
Image Prefetch rules are now more precise, and the monitoring dashboard gives operational teams what they need to understand and resolve failed retrievals themselves.
- Partial body-part matching: a new case-insensitive "contains" option for body part matches all the naming variations hospitals use (for example "THORAX" matches "CT THORAX" and "THORAX EN ABD"), replacing long lists of exact values.
- Most recent study per modality: an optional rule setting fetches only the latest study per modality across all source hospitals, keeping the local PACS free of older, low-value priors. Older studies remain available for manual export.
- Source institution and date of study: two new dashboard columns show which hospital each study came from and the clinical date of the study, both sortable and filterable.
- Failure cause on failed retrievals: failed retrievals now show a clear reason for the failure, with a copy option for support tickets. Technical details and patient data are never exposed.
- Manual retry from the dashboard: a failed job can be retried in one click, with every retry recorded in the audit log.
What this means for you
Clinicians receive the prior studies that are relevant to the encounter, with less noise and less unnecessary data transfer. Clinical and PACS administrators can see why a retrieval failed and resolve it straight from the dashboard, reducing the need to contact support.
Interested? Get in touch with your Founda point of contact to learn more about Image Prefetch and how it can support your organization.
Founda Health Portal: additional patient identifiers and usability improvements
The Patient Details tab of the Founda Health Portal now supports adding and removing additional patient identifiers, matching the legacy viewer. This release also includes a range of usability improvements based on customer feedback.
- Add and remove identifiers: select the assigning authority, enter the identifier and add it; removal asks for confirmation.
- Access controlled and audited: only users with the patient update privilege can make changes, and every change is recorded in the audit trail. Other users see identifiers read-only.
- Usability improvements: the patient list shows the correct number of results, phone numbers are displayed correctly in the patient details, the patient registration and consent widgets open and close as expected, the document upload widget confirms a successful upload instead of continuing to show a loading indicator, and missing interface translations have been added.
What this means for you
Authorised staff can link a patient to identifiers from other domains directly in the portal, supporting patient matching and document retrieval across domains, without switching to the legacy viewer. Everyday tasks such as finding a patient, registering a patient, managing consent and uploading documents work more smoothly and give clearer feedback.
Integration Engine: reading values from JSON responses
The Integration Engine (previously forConnect) now offers a set of functions for reading values from JSON responses, such as the response of an HTTP request, and reusing them in later flow steps.
What this means for you
Integrations with REST APIs can be built with the standard HTTP action, without a dedicated connector just to read a few response fields. New integrations are faster to deliver and can respond reliably to what an external system returns.
What's New – Netherlands
MedMij: imaging records aligned with the Nictiz profile
Imaging records exchanged through MedMij now meet the Nictiz requirements identified during compliance re-testing.
- Standardised codes: document types use SNOMED CT codes from the prescribed value set, LOINC codes use the correct coding system, and anatomy codes are mapped to the Nictiz value set.
- Correct identifiers: practitioner identifiers, study instance UIDs and patient references now carry the required systems and prefixes.
- No impact outside MedMij: the changes are backwards-compatible for other XDS deployments.
What this means for you
Hospitals sharing imaging through MedMij can move forward with Beeldbeschikbaarheid certification, and personal health environments (PGOs) receive imaging records with consistent codes and identifiers.
Defect Fixes
All Markets
- Notification addresses can be reached by subscribers: Fixed an issue in cloud deployments where document notifications (DSUB) contained an internal server address that subscribers could not reach, for example when unsubscribing. The address used in notifications can now be configured to point to an accessible URL.
- Complete trace files for query responses: Fixed an issue where trace files for query responses were cut off at 8192 bytes, making longer responses hard to troubleshoot. Trace files now contain the full response.
- HL7 conversion processes the full message: Resolved an issue where the HL7 conversion action in the Integration Engine did not parse the entire message, so later segments could be missed. The full message is now processed.