To Edit or Not to Edit a Clinical Photo: Where the Line Is

"I just brightened it a bit." "I just removed a shadow." "I just smoothed the shine on the forehead." Almost every manipulation of a clinical photo starts as an innocent improvement. The problem is that the before-and-after photo exists to show a real change, and any edit that alters the look of the face compromises that function, in the chart, in the conversation with the patient and in publicity.

This page separates what is a standardization adjustment from what is manipulation and offers a simple criterion for deciding.

What is the principle?

A clinical photo is a document. It can be processed to be comparable to another, but not to look better. The question that settles most cases is:

Does the edit change what the photo shows about the patient's face?

If the answer is yes, it is manipulation. If it is no (it only makes the photos comparable), it is standardization.

What is acceptable?

See the details in correcting brightness and contrast in a clinical photo, cropping and rotating clinical photos and aligning photos by the eyes.

What is manipulation?

The list includes the most tempting shortcuts.

• Skin smoothing, which erases texture, pores and fine lines.
• Lightening under-eye areas or darkening shadows.
• Erasing marks, spots, scars, veins.
• Reducing swelling or redness through color adjustment.
• Slimming the face, lifting eyebrows, enlarging lips ("liquify" tools).
• Altering only the after photo, with contrast, sharpness or brightness.
• Using filters from phones and beauty apps.
• Using generative AI to "complete" or "improve" the image of an actual result.

Is there a "middle ground"?

The gray zone exists and deserves caution. Two examples:

1. Shine on the forehead. If the light reflected and created a white spot, the solution is to retake the photo, not to "paint" the skin. Reducing the highlight locally is an edit that changes the image.
2. Loose hair. If a strand falls across the forehead, retake the photo instead of erasing the strand by editing.

The practical rule: if the photo needs local editing to be acceptable, it should be retaken. Retaking costs a minute and preserves the document.

How to edit when it is necessary

How do you record the edit?

A simple record avoids future problems. Note, with the file or in the chart:

• the name of the original file and of the edited copy;
• what was changed (for example, "3:4 crop, exposure +0.3 on both");
• the date and who did it;
• the reason (standardization for comparison).

This shows, in case of doubt, that the goal was to compare and not to beautify.

What about publicity?

Publicity is the most sensitive point. Even small adjustments to a result photo are viewed with suspicion, and health advertising rules impose limits. Generally:

• do not publish an edited result photo beyond what is strictly necessary;
• do not use filters or retouching;
• have the patient's authorization (separate consent for publishing images);
• never present a simulation as an actual result.

The role of AI in editing

Generative AI tools can "improve" photos convincingly, and that is exactly why they are a risk: the result looks like a real photo but is a reconstructed image. In a clinical photo, do not use AI to retouch, complete or simulate a result that will be presented as real. Harmy generates an "after" projection as a simulation, clearly marked as such, and does not replace the actual result photo.

What should you tell the patient?

Summary for your routine

Shoot it right so you do not need to edit, keep the original, limit yourself to crop, alignment and equivalent global adjustments, never beautify or alter only one photo, record what you did and respect publicity rules. See also common mistakes in before and after photos.