Live · Available now

DICOM anonymisation
that lives inside
your network.

PixelIQ processes every layer of DICOM identity - header tags, burnt-in pixel overlays, and structured report fields - before any study leaves your PACS. On-prem. Sub-50ms. PS3.15 compliant.

113+
DICOM attributes
<50ms
Per study
PS3.15
E.1 + E.2
Zero
Raw PHI in transit
PixelIQ · live anonymisation stream
(0008,0050)AccessionNumber██████████████stripped
(0010,0010)PatientName██████████stripped
(0010,0020)PatientID█████████stripped
(0010,0030)PatientBirthDate19620000pseudonymised
(0008,0080)InstitutionName███████████████stripped
(0008,0020)StudyDate20240814retained
(0008,0060)ModalityCTretained
(0008,103E)SeriesDescriptionCHEST W CONTRASTretained
(0010,21B0)AdditionalHistory██████████████████stripped
9 of 113 attributes shown · PS3.15 E.1 profile · pixel PHI scan complete
The problem

DICOM carries patient identity
in three places most tools miss two of them.

A standard CT study can expose a patient across header tags, pixel overlays, and structured report fields simultaneously. Header-only de-identification - which is what most PACS export scripts do - leaves two surfaces untouched.

Surface 01
DICOM header fields

113+ PS3.15 attributes - PatientName, DOB, AccessionNumber, InstitutionName, ReferringPhysician and more.

Header scrub · configurable per-tag profile
Surface 02
Burnt-in pixel PHI

Patient banners, accession overlays, clinic watermarks baked into the pixel data. Invisible to header-only tools.

Pixel-level PHI detection · region redaction
Surface 03
Structured reports & PR objects

SR, KO, and GSPS objects carry referenced UIDs, text values, and coded sequences that survive header stripping.

Full SR parse · coded sequence scrub
How it works

The PixelIQ pipeline.

Every step runs inside your network. Nothing raw ever leaves.

01
PACS Intercept
Study arrives via DICOM C-STORE, C-MOVE, or STOW-RS. Minimal network configuration.
02
Pixel Analysis
Pixel-level detection of burnt-in PHI across all SOP classes including secondary capture.
03
Header Scrub
All 113+ PS3.15 attributes processed. SR, KO and GSPS objects parsed and cleaned.
04
Policy Check
Routing rules validated. Recipient allowlist enforced. DUA compliance verified.
05
Governed Output
Clean study forwarded. Cryptographic audit entry written to immutable log.
Integration

Transparent proxy.
Your PACS never knows it is there.

PixelIQ deploys as an on-prem agent and registers as a standard DICOM node inside your network. No changes to your PACS configuration. No radiologist workflow disruption. Outbound-only network model.

DeploymentOn-prem agent · flexible deployment
Network footprintOutbound-only · no inbound connections required
HA / DRHigh-availability · automated failover
MonitoringReal-time metrics dashboard · SIEM integration
Tested PACS / VNA compatibility
Orthanc
DCM4CHEE
Sectra IDS7
Intelerad
Fujifilm Synapse
Agfa Enterprise
Philips Vue PACS
Change Healthcare
RamSoft PowerServer
Mednax / Nighthawk

Any DICOM-compliant PACS supported via standard C-STORE / STOW-RS. Custom connector available on request.

Capabilities

What PixelIQ covers.

Pixel-level PHI detection

Our pixel-level detection model trained on radiology imaging - patient banners, accession overlays, clinic watermarks - identifies and removes PHI that is completely invisible to any header-only tool. Covers all primary imaging modalities and secondary capture objects.

SOP classes covered
CT ImageMR ImageDX ImageMG ImageSC ImageSR DocumentXA ImageNM ImageCR ImageUS Image

PS3.15 header scrub

All E.1 and E.2 profile attributes processed. Configurable retain / remove / pseudonymise per tag, per destination.

SR & KO handling

Structured Reports, Key Objects, and Presentation States processed natively - not skipped as most tools do.

Policy-gated routing

Each destination receives only what its data-use agreement permits. Enforced at the network layer, not in application logic.

Immutable audit chain

Every anonymisation decision is cryptographically signed. Tamper-evident log for HREC, TGA, and OAIC review.

Use cases

When teams reach for PixelIQ.

01

Sending studies to an AI vendor

Your AI vendor needs de-identified DICOM. PixelIQ intercepts studies leaving your PACS, strips PHI on-prem, and forwards clean images - no manual export, no compliance gap.

02

Building a research cohort

Ethics approval covers de-identified images. PixelIQ processes your retrospective archive and routes anonymised studies to a research PACS or S3 bucket - with a signed audit trail for your HREC.

03

Multi-site clinical trials

Standardise de-identification across sites. PixelIQ applies the same PS3.15 profile everywhere, maintains consistent pseudonymous UIDs across the trial, and logs every transfer centrally.

04

Training an internal AI model

Your data science team needs labelled DICOM. PixelIQ de-identifies and routes to your internal GPU cluster - PHI never touches a research environment, governance is automatic.

Technical specifications

Built for PACS integration, not alongside it.

PixelIQ runs as a transparent DICOM proxy. Your PACS does not know it is there. No workflow changes for radiologists. No integration project for IT.

Protocol support
DICOM C-STORE, C-MOVE, WADO-RS, WADO-URI, STOW-RS, DICOMweb
DICOM profile
PS3.15 E.1 Basic Application Confidentiality · E.2 Retain Safe Private Option
Modalities
CT · MR · DX · MG · CR · XA · NM · PT · US · SC and all secondary capture
Pixel PHI detection
Pixel-level detection trained on radiology imaging
Throughput
<50ms per study · async bulk mode
Deployment
On-prem agent · no cloud data path for raw studies
UID handling
Re-map, retain or remove · consistent cross-series mapping maintained
Audit format
Cryptographically-signed audit format · exportable for HREC / TGA review
Compliance
DICOM PS3.15, OAIC APP 11, TGA SaMD pathway, ISO 27001
Live · Available now

See PixelIQ on your DICOM infrastructure.

We will connect to your PACS in a sandbox and run a live anonymisation demo against your study format.