DICOM anonymiser
Free DICOM anonymiser. Remove patient name, ID, dates and hospital details from .dcm files in your browser. See exactly what changed.
What it does
A DICOM file usually carries the patient’s name, identifier, date of birth, the hospital, the referring doctor and the exact date of the scan. If you are about to send a study to anyone, that information travels with it unless you take it out.
This tool removes or replaces those fields following the DICOM confidentiality profile, shows you every field it touched with the value it had before, and re-reads the result to confirm the file is still valid.
Covers the identifying fields
Name, ID, birth date, accession number, institution, physicians, addresses, phone numbers and identifying dates.
Shows exactly what changed
A table of every field, its original value and what replaced it.
Keeps the study intact
Study, series and instance UIDs are deliberately preserved, so the images stay linked as one study.
Verified afterwards
Each output file is parsed again and checked to confirm it is still valid DICOM.
How to use it
Load the files
A single image, a series, or a ZIP of the whole study.
Review what was found
The tool lists the identifying fields that actually carry a value.
Anonymise
Fields are replaced or blanked consistently across every file.
Check and download
Read the table of changes, then save the files or the ZIP.
When people use it
Sharing a case for teaching
Use a real study as an example without exposing a real person.
Sending a scan for a second opinion
Remove details you would rather not pass to a third party.
Submitting images to research
Most protocols require de-identified data before anything is accepted.
Supported files
- Uncompressed DICOM (implicit and explicit VR, little and big endian)
- RLE lossless compression
- JPEG baseline and extended, decoded by your browser
- Files with a .dcm extension, and DICOM files with no extension at all
- ZIP archives containing a whole study or several series
Limitations
- We cannot promise a file is fully anonymous. Private vendor tags may hold copies of identifying data, and text burned into the pixels is not metadata at all.
- Values are rewritten inside the original field, so a replacement that does not fit is blanked instead of truncated. That is safe, but it means some fields end up empty rather than carrying a placeholder.
- JPEG Lossless, JPEG-LS, JPEG 2000 and deflated transfer syntaxes are not decoded — they need heavy codecs we deliberately do not ship. You will get a clear message instead of a blank screen.
- This is a technical tool. It does not read your scan, does not look for findings and does not produce any diagnosis.
- Very large studies of several hundred megabytes may be slow, because all the work happens on your own machine.
Privacy and security
Processed in your browser — nothing is uploaded
Your files never leave your device. Everything runs inside your browser using JavaScript — there is no upload, no server copy and no queue.
Because nothing is uploaded, there is nothing for us to store or delete. Closing the tab removes the files from memory.
We never put file names, patient identifiers or metadata values into analytics or logs. Product events only count how many files were opened.
If you later choose the paid AI explanation, that is a separate, explicit step: you decide which files to attach and nothing is sent until you confirm.
Frequently asked questions
How do I anonymise a DICOM file?
Load it here, review the identifying fields that were found and press anonymise. The cleaned files download straight to your device.
Is the result truly anonymous?
It removes the standard identifying metadata, which is what most sharing requires. It cannot guarantee full anonymity: private tags and text printed into the image may remain. For a formal de-identification, use validated clinical tooling.
What about text printed on the image itself?
Ultrasound, angiography and secondary capture images often have the patient name drawn into the pixels. Metadata removal does not touch that — the tool warns you when the modality is one where this is common.
Will the anonymised files still open?
Yes. Each one is re-parsed and validated after processing, and the study, series and instance identifiers are left alone so the images stay grouped correctly.
Can I anonymise a whole study at once?
Yes. Drop in the ZIP and every file is processed consistently, then returned as a ZIP.