The short version
A ToothPal score is a reading of visible appearance in the photos you took. It is not a clinical measurement, it does not name conditions, and it is not a substitute for a dental examination.
What it is good at: noticing that something looks different this week than it did six weeks ago, in a specific part of your mouth, and putting a number on the size of that difference.
The five dimensions
Each dimension is scored out of 100 against a reference rubric, from a defined photo. They are not interchangeable and they don't all move at the same speed.
| Dimension | What it reads | Photos it uses | How fast it moves |
|---|---|---|---|
| Smile Brightness | Tooth shade against a reference scale | Anterior bite | Weeks to months |
| Gum Appearance | Visible redness or swelling-like patterns along the gum line | Anterior, both buccals, upper and lower arch | Days to weeks |
| Oral Cleanliness | Surface buildup or staining patterns | Anterior, both buccals, upper and lower arch | Days |
| Tongue Cleanliness | Film colour on the tongue surface | Relaxed tongue photo | Hours to days |
| Oral Freshness | Cleanliness and tongue cleanliness, read together | Composite of the above | Days |
Freshness is the only composite: it averages your per-region cleanliness with your tongue reading, which is why it tends to move when either of its parents move.
Why five and not one
A single "oral health score" would be easier to market and much less useful. The five dimensions drift independently and respond to different things.
- Brightness responds to what you drink and how long since a professional clean — it is slow and stubborn.
- Gum appearance responds to interdental cleaning within about two weeks, which makes it the most rewarding dimension to work on first.
- Cleanliness responds within days and is the noisiest reading — it is closest to "what did you do this morning".
- Tongue cleanliness can change overnight, which is why it is the best early indicator that a routine has slipped.
- Freshness is the number most people care about and the least useful to act on directly — fix its parents instead.
How a reading is built
A guided capture flow frames each photo so the same anatomy lands in the same place each time — this matters more than image quality. A shot taken at a different angle under different light is a different measurement, not a worse one.
The model grades each photo against a reference rubric built by a dentist, producing a per-photo score. For Advanced Check-ins, cleanliness and gum appearance are graded per mouth region (anterior, upper arch, lower arch, right buccal, left buccal) rather than averaged into one figure.
The composite Smile Index is a weighted blend, which is why it moves less dramatically than any individual dimension. That damping is intentional.
What a photo cannot see
This is the part most apps skip. A photo of your mouth contains a lot of information and is completely blind to the rest.
- Anything below the gum line or between contacting surfaces — the two places most problems begin.
- Anything inside a tooth. Decay that hasn't reached a visible surface is invisible to a camera and always will be.
- Bone level, root condition, or anything an X-ray exists to show.
- Pain, sensitivity, bite forces, or how any of this feels.
- Whether a change is benign or worth attention — appearance is the signal, interpretation is a professional's job.
If a reading changes sharply and stays changed, that is a reason to book an appointment, not a reason to read more into the number.
Reading your own trend
- Ignore your first score. A baseline is a starting point, not a verdict — the number only becomes meaningful once it has something to be compared to.
- Look at direction before magnitude. A steady +1 a week beats a single +8 that came from better lighting.
- Pick the lowest dimension and work only on that for three weeks. Splitting attention across five is how nothing moves.
- Re-scan on the same day of the week at roughly the same time. Consistency in the measurement beats frequency.
- Expect plateaus. Brightness in particular will flatten and stay flat — that is the dimension doing its job honestly.
Questions
A low score means something appears less than ideal in your photo relative to a reference. It does not mean anything is wrong, and it is not a diagnosis. What matters is the direction it moves over several scans.
Lighting, camera angle, time since brushing and how recently you ate all shift what a photo shows. This is why ToothPal weights trends over single readings and asks you to scan on a consistent schedule.
No, and ToothPal deliberately doesn't let you. Scores are calibrated to a rubric, not to a population, so a cross-person comparison would be meaningless — and being ranked against strangers is exactly the thing that makes people quit.