Image Tools
DICOM Viewer & Anonymizer
View DICOM (.dcm) medical images and all their tags in your browser, adjust the window and level, step through frames and save PNGs. De-identify one file or a whole folder following the DICOM PS3.15 Basic Profile, with options to keep or shift dates. Files never leave your computer.
How to use
- 1Drop one or more DICOM files (with or without the .dcm extension). The first image opens with its own window settings.
- 2To view: drag on the image or pick a preset to change the window, step through frames, filter the tag list, or save the view as PNG.
- 3To anonymise: switch to De-identify, choose what to replace, keep, shift or remove, then download the cleaned file or a ZIP of all of them.
Frequently asked questions
Are my files uploaded?
No. Files are read, displayed and de-identified in your browser; nothing is sent anywhere. That matters for patient data: you can de-identify images on the computer they're on before they go anywhere.
What exactly does De-identify change?
It applies the Basic Application Level Confidentiality Profile from DICOM PS3.15 Annex E: patient name and ID are replaced with what you type; identifying elements (birth date, addresses, physicians, accession number, institution, device serial numbers, comments and more) are removed or emptied; study, series and instance UIDs are replaced with new ones, consistently across all files so the series still fit together; private tags and overlays are removed by default. It records Patient Identity Removed = YES and the profile and options used.
What does it not remove?
Text burned into the image pixels (patient names on ultrasound screens, scanned forms, screenshots) stays, and so do faces in 3-D reconstructions of head scans. The tool warns when a file is marked as having burned-in annotation or comes from a modality where that's common, but you need to look at the images. Check the result against your organisation's rules before sharing.
Which images can it display?
Uncompressed images (little or big endian, implicit or explicit VR, deflated), RLE Lossless, JPEG Lossless and JPEG Baseline, in monochrome (including MONOCHROME1), RGB, YBR and palette colour, single or multi-frame. JPEG 2000, JPEG-LS and video can't be shown, but their tags can be read and the files de-identified.
Can I use it for diagnosis?
No. It isn't a medical device and hasn't been validated for clinical use. Use it to look at, teach with and prepare de-identified data, not to make clinical decisions.
Sources
Window and level in plain words
A CT pixel holds a value in Hounsfield units, from about −1,000 (air) to over +1,000 (bone), far more than a screen's 256 grey levels. The window picks which part of that range is spread across black to white: the level (centre) sets which value is mid-grey and the width sets how many units the grey ramp covers. A narrow brain window (40/80) separates grey and white matter but shows bone as plain white; a wide lung window (−600/1500) shows the airways. Dragging on the image does the same thing interactively.
Example: sharing a scan with a researcher
Drop the whole folder of a CT series (say 240 files). In De-identify, set the patient name to STUDY01 and the ID to S01, choose Shift dates with −90 days so intervals between scans survive, keep sex and age if the study needs them, and download the ZIP. Every file gets the same new Study and Series UIDs, so the series still opens as one stack in any viewer, and the original UIDs, which can be traced back to the hospital, are gone.
What 'consistent UIDs' means
DICOM objects point at each other by UID: a series belongs to a study, a structure set refers to the images it was drawn on. Replacing each UID with a random one would break those links. Here the first time an original UID is seen it gets a new random 2.25.… UID, and every later occurrence of the same original, in any file of the batch, gets the same replacement. De-identify related files in one batch to keep their links.
Related tools
- Aseprite File Exporter (.aseprite to GIF, PNG, Sprite Sheet)
Open an .aseprite or .ase file without Aseprite and export it as a GIF, animated PNG, PNG frames, one folder per layer, or a sprite sheet with Aseprite-style JSON. Tags, hidden layers, groups, blend modes and tilemaps are respected.
- FITS Viewer (Astronomy Images)
Open astronomy FITS images (.fits, .fit, .fts, .fits.gz) in your browser. Stretch them with zscale, percentiles, log, sqrt, asinh or histogram equalisation, read pixel values and RA/Dec under the cursor, browse every HDU and header card, and save the view as PNG.
- Multi-Page TIFF Viewer
Open single and multi-page TIFF files in any browser, including scanned documents and fax TIFFs. Flip through the pages, check each page's size, resolution and compression, and save pages as PNG or the whole file as a PDF.
- PSD Viewer & Layer Extractor
Open Photoshop .psd and .psb files without Photoshop. See the image, show or hide layers, read the text in text layers, and save the image or every layer as PNG, trimmed or full-size, with a JSON list of positions. Nothing is uploaded.
- E-invoice Reader (ZUGFeRD, Factur-X, XRechnung)
Open a ZUGFeRD or Factur-X PDF, or an XRechnung / Peppol XML file, and read it as a normal invoice: seller, buyer, lines, VAT, totals and bank details, with the EN 16931 calculation checks and CSV export.
- PDF Layers Viewer & Remover
See the layers in a PDF (CAD drawings, maps, Illustrator and InDesign files), switch them on and off with a live preview, then delete hidden layers for good or save them switched off.