Case Logbook docs

Help › Photos › Clearing and sharing

Clearing and sharing

In short. Before a photo is shown to anyone but its own patient, you cover identifiers and mark it cleared for showing. Your phone suggests the marks and you confirm each photo. Then you can share a copy, show it in similar cases, or put it in an Album. Nothing leaves your phone unscrambled (where your data lives).

Two things must be true first: the case has consent that covers showing to others, and you have cleared the photo.

Prepare for showing

Tap Prepare for showing on a case page, in a photo’s Edit sheet, on the Teaching file, on an Album, or, with a filter on in Cases → Photos, on the photos you have filtered.

Your phone looks at each photo itself and suggests solid bars over eyes and boxes over printed text such as name labels, hospital numbers and dates of birth. The first time, it downloads its detector to your phone (about 18 MB, once). Detection runs on the phone. No photo, no part of a photo and no result is sent anywhere.

The sheet groups the photos:

  • Suggested marks: check the bars and boxes. Tap a photo to move, resize, add or remove them. Once every photo has been on your screen, Confirm all with these marks confirms the group.
  • Nothing found — check at a glance: the phone did not find anything. That is not a pass. Look at each photo, then confirm it with No identifiers visible, or tap Confirm all once every photo has been on your screen.
  • Not checked by the detector: open each one and check it yourself.
  • Already cleared: open one to change its marks.

You confirm every photo; nothing is ever cleared for you. Cleared does not mean unidentifiable: a tattoo, a scar or a distinctive feature can still identify someone, so judge each photo yourself. A single side marker such as L or R stays visible and outlined; tap it to cover it. If the phone found a face and you confirm without covering it, it asks first. Your marks are saved with the photo, scrambled like the photo itself, and the photo you took is never changed. Mark as not cleared stops a photo being shown to others and keeps its marks for next time.

The case page says how many photos are cleared, for example 4 of 6 photos cleared for showing.

Share a copy

In a photo’s Edit sheet, tap Share a copy. Your phone makes a copy with the marks burned in and every hidden detail removed (no time, place or camera information), and shows it to you. Tap Share to send it with your phone’s share sheet, or Download on a computer.

Share a copy is offered only for a cleared photo of a case with consent that covers showing to others. That holds even if you mean to send it to the patient themselves, because the app cannot know who receives it. Your patient can always see their own photos in Show patient.

Settings → Photos shared lists what you shared and when. It never records who you shared it with.

Albums

An Album is your own hand-picked set of photos from any cases and your Teaching file. In Cases → Photos, tap Select, tick photos or a whole case, and tap Add to album. You can also add a photo from its Edit sheet or the Teaching file. Give each photo a teaching note if you like. Removing a photo from an Album, or deleting an Album, never deletes the photo.

Play shows the Album on screen for teaching: a title page with the Album’s note, then each photo that is ready to show, with your teaching note. The case’s own notes and your Teaching file’s private note never appear. Share makes cleared images (up to 20) or one PDF contact sheet with captions, on your phone, and sends it through the share sheet. Photos that are not ready are left out, and the sheet says how many and why.

Less common

  • Linking a colleague to an Album. There are no web links. Nothing is ever kept on our server in a form anyone could open. To give a colleague an Album, share the contact sheet or the cleared images.
  • Albums and consent. Only photos with consent covering showing to others, and cleared, are ready to show.

Credits

Faces are found with YuNet (OpenCV Zoo, MIT licence) and printed text with PaddleOCR’s PP-OCRv4 text detector (Apache-2.0 licence), both running on your phone on ONNX Runtime (MIT licence), the same engine the app uses for voice.