Turn a script your clinical team has already approved into a short presenter-led clip that explains a procedure step by step. Add a voice track, simple visuals and captions, then publish a version in each language your patients speak. A qualified professional signs off on every word.
A patient who understands a procedure arrives calmer and asks better questions, and most clinics already know it. The bottleneck is time: nobody can re-explain a root canal or a bowel prep to every patient, every single day. A short clip can do the repeating for you. This page shows how to build one with a digital presenter, a voice track and simple visuals, and where a qualified professional has to stay in the loop.
Most patient handouts are written once, in clinical language, and read by almost nobody. A two-minute video that walks through what happens before, during and after a procedure is easier to follow, easier to replay at home, and easier to share with a worried relative who could not come to the appointment. Until recently the obstacle was production: a camera, a quiet room, a staff member willing to be filmed, and an editor.
Picasso IA removes most of that. A presenter model such as Avatar IV reads a written script on screen with a digital avatar and voice, so there is nothing to film. Because the whole thing starts from text, updating a clip after a protocol changes means editing a paragraph and generating again, not rebooking a shoot. And once a version exists, the same script can be voiced or translated for patients who do not speak the clinic's main language. The toolkit is where all of it happens, and new accounts get free credits to try a first clip.
Not every explanation needs a presenter. A clip about how long to fast before a procedure is fine as a spoken checklist, while a consent conversation is better left to a human altogether. Match the format to what the patient has to do or understand.
| Clip type | Good for | How to build it | Watch for |
|---|---|---|---|
| Presenter clip | Pre-visit overview, what to expect on the day | Avatar IV reads the approved script on screen | The presenter is digital, not your clinician, so say so |
| Narrated visuals | Aftercare steps, simple sequences | Images from nano-banana-pro, voice from speech-2.8-hd | Drawn anatomy can be wrong, so keep visuals simple |
| Animated still | A single moving diagram or icon | An image animated with wan-2.6-i2v | Motion can distort shapes, review every second |
| Captioned loop | Waiting-room screen, no sound | Short clip with captions switched on | Text must stay readable at a distance |
| Translated version | Patients who speak another language | video-translate on a finished clip | A bilingual staff member checks the result |
Whatever the type, keep one clip to one question. How to prepare for a cleaning and what to do after a tooth extraction should never share a video, because patients replay the part they need and skip the rest.
This walkthrough assumes you already have a script that a dentist, nurse or physician has written or approved. The tools only turn those words into a clip.
The script matters more than any model you choose. Patients are often nervous, in a hurry or reading in their second language, so write the way you would speak to someone at the front desk, in short sentences with one idea each. Open with what the clip covers and how long it takes, then go in the order a patient lives it: before, during, after.
Name things in everyday words first and add the clinical term once, after the plain version. Give instructions as actions, such as avoid hot drinks until the numbness has gone, rather than as restrictions. End by saying who to call with questions, using the clinic's own contact details, and keep dates, doses and times out of the voice track unless your clinician has confirmed them for that patient. A simple outline works well: what it is, why it is done, what you will feel, what to do afterwards, when to call us.
Read the script aloud before generating anything. If you stumble, so will the voice. The voiceover page covers pacing and pronunciation in more depth, and the explainer videos page shows how the same structure works outside healthcare.
An AI presenter reads whatever you give it with total confidence, and that is exactly why health content needs a human gate. These three points are not optional.
Medical accuracy has to come from a qualified professional. The model does not know your patient, your protocol or your jurisdiction, and a clip it voices is general information, not medical advice.
First, every claim, step and warning in the script must be written or approved by a qualified clinician, and that same person should watch the final clip. Second, label the video as general information about a type of procedure, not advice about an individual patient, and direct people to the clinic for their own situation. Third, if the clip is used to promote services rather than to inform existing patients, health advertising rules apply, and compliance with them is the clinic's responsibility. Those rules differ by country and profession, so check the current requirements of your regulator and platform rather than relying on this page.
On faces and voices, use fictional presenters or a presenter you have the right to use. Never build a clip around a real, identifiable person who has not agreed, and never present a digital avatar as your actual doctor.
This is a tool for explaining, and there are places where it stops being useful or becomes risky.
If a clip would need heavy editing to be safe, a short written handout or a call from the nurse is the better tool.
Yes, as a production tool. On Picasso IA a presenter model reads a script on screen with a voice, and you can add simple visuals and captions around it. The content itself still has to come from a qualified professional, and one should review the finished clip. Treat the model as a camera crew that never gets tired, not as a source of medical knowledge.
Avatar IV is the one built for this: you give it a script of up to 5000 characters, pick an avatar and a voice, and choose a display style such as full frame or close-up. It also offers voice emotions like Soothing and Friendly, which suit a calm explanation. Try a short test script first to see whether the avatar and voice feel right for your patients.
No, and it should never be presented as such. A generic clip explains how a type of procedure usually works. It cannot take your patient's history into account, so label it general information and tell viewers to ask the clinic about their own situation. Anything that sounds like a personal recommendation belongs in a consultation with a qualified professional, not in a pre-recorded video.
The clinic and its qualified professionals. The model has no clinical judgement and will read an error as fluently as a fact. Have a dentist, nurse or physician write or approve the script, then watch the finished clip with captions on, because visual details and subtitles can introduce mistakes the script did not have. Keep a record of who approved which version.
You can, but health advertising rules then apply, and following them is the clinic's responsibility. Those rules vary by country, by profession and by platform, and they often cover claims about results, before and after material and testimonials. Check the current requirements from your regulator and from the platform where the video will run, and ask your compliance contact when in doubt.
Yes. You can voice the script in other languages, or run a finished clip through video-translate, which dubs a video into many languages and syncs the lips to the new audio. Translation can shift tone or lose a caveat, so someone fluent in the target language should review each version. Medical terms are the usual trouble spots, so check those first.
Only with their clear permission, and only with likenesses and voices you own or are cleared to use. Do not build a clip around a real person who has not agreed, and do not present a digital avatar as an actual clinician at your practice. Many clinics simply use a clearly digital presenter and state it, which keeps expectations honest.
Use icons, outlines and plain illustrations instead of realistic anatomy, and keep one idea per scene. Generate them with an image model like nano-banana-pro, then look closely for odd shapes, extra parts and unreadable labels. If a visual explains something clinical, ask your professional to confirm it. When in doubt, drop the picture and let the presenter and captions do the work.
Short enough that a patient finishes it. Many clinics aim for one to three minutes per question, though there is no rule, and your own patients will tell you by where they stop watching. Split a long topic into separate clips, one for preparation, one for the visit and one for aftercare, so people can replay only the part they need.
Generations are paid in credits, and the cost depends on the model and the settings you choose, so any number written here would go stale. A presenter clip, a voice track and an animated image are each priced separately. New accounts get free credits to test a first clip, and the current plans are on the pricing page.
Ready to try it? Open the Picasso IA toolkit, paste a script your clinician has approved and generate a first clip to review.
Choose your language
