Radiology read de-identification is the removal of patient identifiers from the file. UK GDPR Art. 9 and the DPA 2018 govern health data. anonym.plus runs this on your own machine. The findings stay readable, but the text no longer names a person.
When this applies
You want to teach from a chest CT write-up or share it with a peer. First the name, the dates, and the order number have to come out.
How anonym.plus handles it
- Open the file (PDF, DOCX, or scan) on your device.
- Local OCR reads scanned pages, so printed text is caught too.
- The tool marks names, dates, MRNs, and phone numbers.
- Review each mark and keep any clinical term flagged by mistake.
- Replace each ID with a steady label, or black it out.
- Save the clean file. The first copy never leaves you.
What you need to provide
- The file (PDF, DOCX, TXT, or image scan).
- An operator: Replace (swap), Redact (remove), or Mask (partial).
- Optional: a name map to re-link cases later.
Patient data entity types detected
| Category | anonym.plus entity type | Example |
|---|---|---|
| Names | PERSON | James Okafor → [PATIENT_1] |
| Dates | DATE_TIME | Study 19/02/2026 → [DATE] |
| Record IDs | MEDICAL_RECORD_NUMBER | MRN 553901 → [MRN] |
| Order | ID | Accession R-44820 → [ACCESSION] |
| Contact | PHONE_NUMBER | +44 113 496 0144 → [PHONE] |
| Location | LOCATION | 88 Birch Avenue, Leeds → [ADDRESS] |
Compliance achieved
- Anonymisation of the free-text read under UK GDPR Art. 9 & DPA 2018; health data is special-category data needing a specific processing condition.
- Holding the read for care usually relies on DPA 2018 Sch. 1, para 2 (health or social care purposes) — the anonymisation step below is separate from that lawful basis; it is about who else gets to see a copy.
- Once truly anonymous, the file leaves UK GDPR scope under Recital 26.
- Assessed against the ICO motivated-intruder test (ICO Anonymisation Code of Practice) — a chest CT read plus a rare finding can still be re-identifying in a small department.
- Working files are kept safe with AES-256-GCM.
Anonymise radiology reads offline — see plans & start free →
Limitations & cautions
UK GDPR Recital 26 asks whether a motivated intruder could re-identify the person. The tool removes the listed identifier types. A rare finding plus a small town can still hint at a person. Review the output before release.
Frequently asked questions
Will the impression still make sense after the swap?
Yes. The Replace operator drops a steady label in place of each ID. The findings and the impression read as before and point to no real person.
Does this read the radiologist's signature block too?
Yes. Names, phone numbers, and dates in the footer are flagged like any other field. You decide whether to keep the reading doctor or hide them.
Can I process a scanned fax of the write-up?
Yes. Local OCR pulls the text from the scan first. Then the same flags apply, so a faxed page is handled like a born-digital one.