NextRIS
IA-native Radiology Information System
Radiology system + PACS. Management of DICOM medical imaging and radiology information for your institution.
NextRIS: the IA-native RIS
NextRIS is born IA-native — not a legacy RIS with AI bolted on the end. The architecture integrates clinical reasoning at every stage of the radiology workflow, with a data layer that normalizes HL7, DICOM and SNOMED CT independent of source.
Why NextRIS
Traditional RIS solve scheduling and worklist. NextRIS solves the full workflow: from an AI-motivated order to a structured DICOM SR report distributed by channel and profile. The difference is measured in hours of reduced reporting time and critical findings automatically prioritized.
Architecture
The system has three layers:
- Data layer: normalizes HL7/DICOM/SNOMED CT, PostgreSQL schema with
nextris.prefix, 50+ relational tables. - Model layer: Nexi (Claude) for clinical reasoning, Whisper for dictation, in-house medical NLP for terminology mapping.
- Orchestration layer: REST API with 200+ endpoints that calls models at the right workflow moment, auditing every decision.
Open by design
We use open standards (DICOM, HL7, SNOMED CT, FHIR-ready) so NextRIS coexists with your existing PACS/HIS ecosystem without lock-in. The integrated viewer is OHIF over DICOMweb with Keycloak SSO.
Related articles
DICOM SR and the structured report radiology needs
Free-text radiology reports lose data at every integration. DICOM SR solves this by turning dictation into traceable structure.
Whisper for radiology dictation: transcription with medical context
Whisper transcribes voice to text, but radiology dictation needs clinical vocabulary, finding names and tolerance for jargon. Here's how we adapt it.
DICOM SR and the Structured Report Modern Radiology Needs
Free-text radiology reports lose valuable information at every integration point. DICOM SR solves this by transforming dictation into structured, traceable clinical data.