Pathology Practices

The slide is de-identified.
The label is not.

Paige.AI and PathAI want your SVS, NDPI, and MRXS files. The scan is clean. The embedded label image - the one with the printed patient name and barcode - is not. SlideIQ handles all three PHI surfaces before any file reaches an AI vendor or research archive.

WSI
TIFF / NDPI / SVS
LIMS
Integration
PS3.15
Aligned
SlideIQ · WSI metadata stream
ImageDescription████████████████████████████stripped
SoftwareLeica Aperio v12.4.3retained
Artist████████████████████stripped
DateTime2024:08:14 00:00:00pseudonymised
Label imageRedacted automaticallystripped
LIMS CaseID█████████████████████stripped
StainProtocolH&E standardretained
Objective40xretained
8 fields shown · TIFF / NDPI / SVS supported62ms
What changes

Before and after SlideIQ.

Everything between your lab and a Paige.AI or PathAI data-sharing agreement.

Without SECUVA
Paige.AI wants SVS files. Legal flags that patient names are in the embedded label
With SlideIQ
SlideIQ strips metadata, label image, and LIMS link. Paige.AI gets clean slides.
Without SECUVA
Lab assistant manually checks slide labels before each export - time-consuming and inconsistent
With SlideIQ
Automated label detection redacts PHI on every slide, every time
Without SECUVA
LIMS case ID visible in exported filenames and metadata
With SlideIQ
Pseudonymous reference replaces case ID - relinkable under controlled conditions
Without SECUVA
No audit trail for which slides were sent to which researcher or vendor
With SlideIQ
Immutable export log with processing provenance for HREC, NATA, and institutional review
Without SECUVA
AI partnership delayed 3+ months while legal reviews data-sharing agreements
With SlideIQ
PHI-free pipeline compresses legal review. Partners go live faster.
The overlooked surface

Everyone de-identifies the scan.
Nobody de-identifies the label.

Whole-slide images carry patient identity across three distinct surfaces: the TIFF/NDPI metadata, the embedded slide label (often a barcode with printed name), and the linked LIMS case record. Standard export pipelines address at most one. SlideIQ addresses all three.

Typical WSI export coverage
TIFF metadata fieldscovered
Embedded label imagemissed
LIMS case linkagemissed
Slide barcode textmissed
Associated lab notesmissed
How it works

The SlideIQ pipeline.

01
WSI Ingest
TIFF, NDPI, SVS received from scanner or LIMS
02
Label Redaction
Automated detection removes label image PHI
03
Metadata Scrub
All TIFF / vendor metadata fields processed
04
LIMS Unlinking
Case ID replaced with pseudonymous reference
05
Governed Output
Clean WSI delivered. Audit entry written.
Day one

Three things that change immediately.

Deployed in your lab environment

We connect SlideIQ to your scanner output and LIMS. Half-day installation. No disruption to the scan-to-report workflow your team already runs.

Typically 4–6 hours on-site

Three-surface coverage active

Metadata, embedded label images, and LIMS case IDs are all processed from the first slide. Paige.AI or your research partner receives governed data immediately.

SVS · NDPI · MRXS · TIFF all supported

HREC-ready export log

Every WSI movement is logged with processing provenance - methodology version, surfaces addressed, audit hash. Ready for HREC submission or institutional review from day one.

Exportable manifest on every release
Capabilities

What SlideIQ covers.

Three-surface de-identification

SlideIQ is the only system that addresses all three PHI surfaces in a whole-slide image in a single pass: TIFF/NDPI metadata, embedded slide label images (including barcodes), and linked LIMS case records. Each surface uses a different technique - metadata scrub, automated label redaction, and pseudonymous key replacement.

File formats
TIFFBigTIFFNDPISVSMRXSSCNLB

Label image redaction

Automated detection identifies slide labels, barcodes, and printed patient text embedded in the slide image tile.

LIMS case unlinking

Case IDs replaced with pseudonymous references. Linkage maintained for re-identification under controlled conditions.

Policy-gated routing

Each AI vendor or research archive receives only the fields and studies permitted by their data-use agreement.

Immutable audit chain

Every de-identification decision logged. Exportable for HREC, TGA, or institutional review.

Technical specifications

Built for the formats your lab actually uses.

SlideIQ integrates with your existing LIMS and scanner workflow. No changes to how your lab operates.

WSI formats
TIFF, BigTIFF, NDPI (Hamamatsu), SVS (Aperio), MRXS (3DHISTECH)
Label redaction
Automated detection - printed text, barcodes, QR codes, handwritten annotations
Metadata scrub
TIFF tags, vendor private tags, XMP/IPTC metadata, embedded XML
LIMS integration
HL7 v2, FHIR R4, direct DB connector (configurable)
Throughput
Real-time · async tile processing
Deployment
On-prem agent · no patient data in cloud pipeline
Audit format
Cryptographically-signed audit format · HREC-exportable
Compliance
OAIC APP 11, TGA SaMD, ISO 27001 · NATA-lab compatible
Questions

What pathology teams usually ask us.

01

Paige.AI wants our whole-slide images. What is the actual privacy risk?

The risk is the label, not the tile. The TIFF pixel data - the scan itself - carries no identity. But every WSI has an embedded label image (often a photograph of the physical slide with a printed patient name and barcode), plus TIFF metadata and a LIMS case ID linkage. Paige.AI, PathAI, and similar vendors will receive all of that unless something processes it first. SlideIQ handles all three surfaces.

02

We already de-identify TIFF metadata before export. Isn't that enough?

Metadata de-identification covers the TIFF tag fields - ImageDescription, DateTime, Artist, and similar. It does not address the embedded label tile (a separate image embedded in the file), the LIMS case ID in the filename or associated metadata, or slide barcode text that may be captured in the label photograph. These are the surfaces that standard export tools miss and that most researchers and AI vendors inadvertently receive.

03

Does SlideIQ work with our LIMS system?

Yes. SlideIQ integrates with HL7 v2, FHIR R4, and direct database connectors to handle LIMS case delinking. Case IDs are replaced with pseudonymous references. The mapping is retained under controlled conditions so slides can be re-identified if required - for example, for a clinical follow-up after a research finding.

04

What about the barcode on the physical slide label?

SlideIQ's label detection model is trained specifically on slide label formats - printed barcodes, printed patient names, handwritten annotations, and QR codes embedded in label images. It detects and redacts the label tile within the WSI file. The WSI tile data (the scan itself) is untouched.

05

How do we satisfy HREC requirements for pathology data sharing?

HREC approval for pathology data sharing typically requires documented evidence of de-identification methodology, an audit trail, and alignment with OAIC APP 11. SlideIQ produces a signed processing manifest for every file - methodology, processing version, surfaces addressed, and audit hash - exportable in the format your institution's research governance team requires.

06

What is the throughput for a WSI batch through the pipeline?

Processing time depends on file size and label complexity. A typical SVS or NDPI file at 40x runs 2–5 seconds. For large batch exports - moving a cohort of slides to a research archive - SlideIQ queues and processes in the background with no impact on routine lab operations.

Get started

See SlideIQ on your whole-slide images.

We will run a live demo against your WSI format and LIMS connection. No data leaves your site.

No data leaves your site during the demo. Half-day on-site engagement.