AI System for Tracking Dental Laboratory Revisions

The problem

Staff manually compare correction notes and scanned job sheets submitted by clinics with the original job card.

When repeat requests and additional changes are not tracked separately, technicians have to ask which revision to prepare.

Production, return dispatch and clinic receipt are recorded separately, so finding the latest version’s status requires a call.

What we set out to improve

• Whether each request links to the correct laboratory job code and previous revision
• Whether duplicate requests can be distinguished from additional changes
• Whether approved scope, return dispatch and clinic receipt can be found in the same revision record

How the system works

1. 01Receive the form from an authorised clinic account with the laboratory job code and revision reference.
2. 02Separate documents linked to patient identity into restricted storage. Show only the job code in the administrative queue.
3. 03OCR converts writing in a scanned document into text. In an authorised reading environment, extract proposed revision details and the requested delivery date from the job sheet and note.
4. 04Link the request to the previous revision. Attach repeat submissions to the existing record and show additional changes separately.
5. 05Prepare a revision job card and a list of open questions for the technician. Present extracted fields alongside their source notes.
6. 06The responsible person approves the scope. Track production, return dispatch and the clinic’s receipt confirmation in the same revision record.

Methods we used

• Match requests through the authorised form, clinic record and laboratory job code.
• Limit document reading and text extraction to revision details and the requested delivery date.
• Organise the revision sequence using previous revision links and duplicate checks.
• Keep the scope decision, production status, return dispatch and receipt confirmation in a shared job card.

Where people stay involved

Data and security

Health information is kept separate from general storage and general-purpose document reading. Using a job code does not make data anonymous. Processing conditions, access, transfers, retention and deletion are assessed with the organisation. Staff review job sheets until an authorised reading environment is available. Health content is received through the authorised form, not WhatsApp. Links between source documents, extracted fields and authorised corrections are preserved.

Who this suits