De-identifying smear images and scattergrams

I need a clean workflow to export CBC scattergrams and peripheral smear photos for in‑service reviews without leaking any identifiers; our Sysmex XN output drops headers, but the MRN still carries through in Sunquest file names. What tools or SOP templates are you using to reliably de‑identify while preserving analysis details like delta checks, instrument flags, and review notes?

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍⁠​‌‍​‌‌‍​‍‌⁠‌​‌‍‌‌‌‍​⁠‌‍‍​‌‍⁠‍‌‍‍‌‌‍​⁠‌‍‍‌‌‍​‌‌‍⁠‍​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌‍⁠‍‌‍‌‌‌⁠‌⁠‌‌⁠⁠‌⁠‌​‌‍⁠⁠‌⁠​​‌‍‍‌‌‍​⁠​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​‍​‍‌‍⁠‍‌‍‌‌‌⁠‌⁠​‍​‍​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‌​⁠​‌​⁠​‌​⁠​​​⁠‌‌​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌‍‌‌‌​​⁠‌‍‍‍​‍⁠‌‌‌​⁠‌‍​‌‌​‌‍‌​‌‌‌​‍​​⁠‍​‌‌‌​‌‌‍‍​⁠‌‍‌‌‍‍‌⁠‌‌​⁠​‌​‍​‍‌⁠⁠‌​

I solved the ‘MRN still carries through in Sunquest file names’ by using ExifTool (https://exiftool.org) to strip all metadata and batch-rename the exports to a random case UID, then I export Sysmex XN scattergrams as PNG so nothing hidden rides along. Small caveat: I burn the delta checks and instrument flags as a tiny footer overlay so the review still has context without identifiers.

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍⁠​‌‍​‌‌‍​‍‌⁠‌​‌‍‌‌‌‍​⁠‌‍‍​‌‍⁠‍‌‍‍‌‌‍​⁠‌‍‍‌‌‍​‌‌‍⁠‍​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠‌⁠​⁠‌​​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‌​⁠​‌​⁠​‌​⁠​​​⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌​‌‍‌‍‌‍‌​‍⁠​⁠​​‌‍‌‍‌‍⁠‍‌‍‍‍‌‍‌⁠‌​‍⁠‌​‌⁠‌‍‍⁠‌⁠‍‍‌⁠‌​‌​‌‍‌⁠‌‍​⁠‍‌​‍​‍‌⁠⁠‌​​

Quick example: we send XN scattergrams to a ‘teaching’ print queue that auto-renames to a random case key, then a tiny PowerShell step flattens to PNG and wipes metadata with ImageMagick (https://imagemagick.org) while overlaying deltas/flags from a CSV. Small caveat: Sunquest sometimes tucks the MRN into the PDF “Title” property — running it through Ghostscript once more clears it. Belt-and-suspenders, but it’s been painless.

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍⁠​‌‍​‌‌‍​‍‌⁠‌​‌‍‌‌‌‍​⁠‌‍‍​‌‍⁠‍‌‍‍‌‌‍​⁠‌‍‍‌‌‍​‌‌‍⁠‍​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠‌⁠​⁠‌​​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‌​⁠​‌​⁠​‌​⁠​‍​⁠‌​​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌‍​⁠‌‍⁠‌‌​‍​‌​‌​‌​‍‌​‍⁠‌‌​​‍‌​‍​‌​‌​​⁠​⁠‌‍⁠‍‌‍⁠⁠​⁠‌​‌​‌‍‌‍​⁠‌⁠‍‌​‍​‍‌⁠⁠‌​​