Dental · Treatment planning

A manufacturable plan in one day.

Planning a clear-aligner case normally takes three to four weeks, and increasingly assumes accelerated compute. This does it in a day, on general-purpose CPUs.

The pipeline

Six stages, and the order matters.

Each stage constrains the next, and the clinician sits inside the loop rather than waiting at the end of it.

  1. Intake

    The scan is checked for coverage, artefacts and occlusal registration. A case that fails comes straight back rather than three weeks later, which is where a good deal of conventional rework starts life.

  2. Segmentation

    Teeth separated from the arch mesh, with roots estimated where the scan does not carry them. Without root estimation the staged movement is a surface animation rather than something physically plausible.

  3. Target setup

    Final position derived from the prescription and the anatomy of the case: arch form, available space, midline, and the occlusal relationship being aimed at.

  4. Staging

    Movement broken into steps that stay inside per-tooth limits. This is where a plan becomes manufacturable or does not. The limits are biological, and a sequence that ignores them produces aligners that will not track.

  5. Attachments and interproximal reduction

    Placed against the staged movement, so the mechanics each stage needs are actually available.

  6. Review and release

    The clinician reviews, adjusts where needed, and releases to manufacture. A revision re-enters at the stage it affects instead of restarting the case.

Five-axis milling of the appliance from the released plan.

The software

Eleven views of one case.

This is the planning environment the pipeline runs in, on case CA90-09928. Scroll through it.

Front view of the case as scanned, upper and lower in occlusion.
01 / The caseFront view of the case as scanned, upper and lower in occlusion.
Left buccal view. Every view is the same model, not a rendered image.
02 / LeftLeft buccal view. Every view is the same model, not a rendered image.
Upper arch from the occlusal, with the existing restorations visible.
03 / UpperUpper arch from the occlusal, with the existing restorations visible.
Back view, showing the vertical relationship between the arches.
04 / BackBack view, showing the vertical relationship between the arches.
Right buccal view.
05 / RightRight buccal view.
Lower arch from the occlusal.
06 / LowerLower arch from the occlusal.
Upper arch hidden, so the lower can be assessed on its own.
07 / Isolate an archUpper arch hidden, so the lower can be assessed on its own.
Measurement grid on, with the aligners shown over both arches.
08 / Aligners in placeMeasurement grid on, with the aligners shown over both arches.
Upper aligner alone against the grid, seated over the crowns.
09 / One alignerUpper aligner alone against the grid, seated over the crowns.
Interproximal reduction called on the plan: 0.2 mm, at stage 4.
10 / Interproximal reductionInterproximal reduction called on the plan: 0.2 mm, at stage 4.
The same call from the occlusal, against the grid.
11 / Checked against the gridThe same call from the occlusal, against the grid.

The part you can check yourself

Send a case and time it.

Of everything on this website, this is the claim a customer can test in an afternoon without a contract, a trial protocol or a word from us.

A manufacturable plan in a single day, on general-purpose CPU hardware.

Input
Intraoral or model scan, with the prescription you would normally give a technician.
Output
A staged, manufacturable plan with attachments and interproximal reduction, released after clinician review.
Turnaround
One day, against an industry norm of three to four weeks.
Compute
General-purpose server hardware. No accelerator required at any stage of planning.
Deployment
Hosted or on-premise. On-premise is normal for laboratories handling their own patient data.
Evaluation
Send an anonymised case you already know the answer to, and compare.

The measure worth watching during an evaluation is not whether the plan looks acceptable. It is the acceptance rate without modification, across a representative spread of case complexity. A pipeline that returns quickly and then needs reworking has moved the queue, not removed it.

Send a case.

An anonymised scan and a prescription are enough to start.

Contact us