Clinician-led quality validation for treatment-planning models
Overview
A dental and orthodontic company building CBCT-scan segmentation models for clear-aligner and orthodontic treatment planning engaged us to build a clinical QA pipeline that validates its models' outputs.
For every patient case, three model outputs were validated: crown segmentation, axes setup, and CT segmentation of bone and teeth. Each one went through consensus review by three expert orthodontists.
How the review ran
Three model outputs, one review team
The team validated three distinct outputs from the client's pipeline: crown segmentation, axes setup (reviewed per jaw, with upper and lower scored separately) and CT segmentation (bone and teeth annotations combined into a single review per case).
Consensus labelling with expert orthodontists
Three expert orthodontists independently reviewed the same task, and their judgments were reconciled into a single consensus label. This reduced the effect of individual clinical subjectivity on the delivered data and gave the client a multi-expert rationale wherever opinions diverged.
A structured anomaly taxonomy
Working with the client's QA and research team, our clinicians applied and refined a shared anomaly taxonomy covering missing teeth and each tooth's identity, position and anomalies, standardised throughout on the FDI tooth-numbering system.
How it stayed aligned
Benchmarked against pre-reviewed cases
Each delivery was scored against the client's own pre-reviewed benchmark cases.
A clinician-to-clinician clarification loop
Our orthodontists raised edge-case questions directly with the client's clinical team and turned every answer into updated internal review guidelines, keeping the two teams' clinical judgment aligned.
How we keep an evaluation fair is set out in our independence policy. See also who runs Biz-Tech.