DICOMUnity

A new foundation for the next era of medical imaging.

AI-ready, with patient information protected. Ready for the quantum paradigm. Clean, modern architecture. Secure by design—from the first line of code to deployment.

For more than three decades, medical imaging has evolved through DICOM, the internet, web viewing, mobile devices, cloud infrastructure, and now AI.

Many systems still depend on architectures created for a different era.

DICOMUnity is a clean, modern medical imaging foundation designed around three principles: AI-ready workflows with protected patient information, preparation for the opportunities and security threats of quantum computing, and secure-by-design engineering from the first line of code to deployment.

DICOMUnity is the bedrock of a new generation of enterprise imaging and health-information systems. PACSFolio turns it into a modern archive. DICOMAgentRouter moves studies between sites through end-to-end post-quantum protection. DICOMAgent brings AI-ready viewing to the radiologist. The DICOM Application Firewall inspects and protects every DICOM exchange. RISArella connects scheduling, reporting, and clinical workflow. DICOMBrain keeps model management, fine-tuning, RAG, and inference on-premise—where patient data remains under institutional control.

Modular systems. One shared foundation. From the first image received to the intelligence that acts on it.

Explore the product ecosystem
01
THE DICOM FOUNDATION LIBRARY

A Rust-native DICOM implementation derived directly from the standard.

Built from the standard. Not from legacy.

Every product shares a clean-room DICOM implementation derived directly from NEMA PS3.x and built from the ground up in pure Rust. It provides a complete data, network, codec, and clinical-object stack without requiring a C/C++ runtime for core operations.

Memory-safe concurrency Memory safety without a garbage collector, with an async-first architecture built for concurrent workloads.
Predictable performance profile No interpreter overhead, no JIT warmup, and no memory pressure from a managed heap.
Complete DICOM data engineExplicit and Implicit VR, transfer syntaxes, character sets, and the DICOM JSON Model. Full network and DIMSE stackAssociation negotiation plus C-ECHO, C-FIND, C-STORE, C-MOVE, C-GET, and N-services. Production imaging codecsJPEG, JPEG-LS, JPEG 2000, RLE, and uncompressed pixel data. Advanced clinical objectsStructured Reports, Presentation States, and encapsulated documents.
Read about the foundation
02
QUANTUM-SAFE BY DEFAULT

Post-quantum DICOM transport without protocol changes.

End-to-end protection—without breaking DICOM.

Protected health information must remain confidential for decades. Encrypted traffic can be captured today and stored until quantum computers are capable of breaking current public-key protection—a strategy known as “harvest now, decrypt later.” Long-lived medical data needs post-quantum protection before that threat arrives.

We designed a transport architecture that makes DICOM networking more secure and quantum-safe without changing the protocol. It creates a secure site-to-site tunnel over standard WebSocket, multiplexes concurrent associations without modifying the PACS, and encrypts every DICOM PDU between endpoints while preserving C-FIND, C-STORE, C-MOVE, C-GET, SOP classes, and transfer syntaxes. Post-quantum protection follows NIST-standardized cryptography.

Site-to-site transfer—no VPN Concurrency—without touching the PACS End-to-end encryption—without breaking DICOM
Explore secure transport
03
AI-READY

On-premise AI under institutional control.

Fine-tune AI without moving patient data.

Authorized users can use DICOMAgent tools to converse with DICOM studies and the institutional archive in natural language—finding relevant cases, exploring imaging context, retrieving similar studies, and drafting reports through locally hosted AI.

Patient images, reports, prompts, and adjusted model parameters remain inside the facility. The institution controls the data, the workflow, and the resulting intelligence.

  • On-premise fine-tuning and model serving
  • Local retrieval grounded in the institutional archive
  • Air-gap capable after the initial model download
Read about on-premise AI
Where we create value

For imaging organizations ready to move beyond incremental change.

Different organizations begin from different points: a PACS that must be modernized, a clinical workflow ready for AI, or an archive platform preparing to scale. DICOMUnity provides a standards-native path forward without discarding the systems and knowledge already in place.

01 / PACS MODERNIZATION

Keep the PACS.
Upgrade what surrounds it.

For hospitals and imaging networks that already operate a PACS but need stronger inter-site protection, cleaner integration, better resilience, and a practical path toward the post-quantum era.

Explore quantum readiness
02 / AI IN THE CLINICAL LOOP

Move AI from an experiment into the imaging workflow.

For radiology teams that want authorized users to search, retrieve, triage, compare, and draft with AI while patient information, prompts, models, and results remain under institutional control.

Explore AI readiness
03 / VNA & IMAGING PLATFORMS

Operate a standards-native archive built to grow.

For VNA providers, regional imaging networks, and archive operators that need multi-vendor ingestion, secure site connectivity, scalable storage, auditability, and dependable DICOM interoperability. Post-quantum protection is put in place before the threat arrives.

Explore archive infrastructure