How to scan a full arch for an implant bridge

What do you need before you scan a full arch?

Before you scan a full arch, have three things settled. The tooth and bite reference, usually the provisional the patient is wearing and you have approved. The abutments, with brand, platform, cuff height and angle of each one noted for the order. And a scan body from the right system for every abutment.

The provisional is the prototype of the definitive bridge: it holds the occlusion and vertical dimension that already work in the mouth. An and colleagues (The Journal of Prosthetic Dentistry, 2018) describe a digital technique that carries that information from the provisional to the definitive without a stone cast. That’s why it gets scanned twice, in the mouth and out of it.

Use the scan body as it came out of the package, with no wear or trimming. Petchmedyai and Thanasrisuebwong (PLOS One, 2023) advise against modifying scan bodies, because the CAD software may stop matching them to its library. The model matters too: in an in vitro study by Meneghetti and colleagues (PLOS One, 2023), with six implants on multi-unit abutments, seven scan bodies and four scanners produced different accuracy.

Why does the scan body decide passive fit?

The scan body decides passive fit because the software calculates each implant position from its surface in the file. In a full-arch bridge, every one of those positions goes into a single rigid framework: one implant off in the file becomes a bar or a zirconia bridge that seats under strain.

The margin is small. Pozzi and colleagues’ systematic review (Clinical Implant Dentistry and Related Research, 2025) treats a misfit under 150 µm and up to 1° of angular deviation as generally enough to avoid complications like screw loosening. In some of the in vitro studies in that review, mean angular deviation with intraoral scanning went past 1°.

Two in vitro studies show where the error comes from. Petchmedyai and Thanasrisuebwong (PLOS One, 2023) exposed the scan body at 8, 5.3 and 2.7 mm and deleted part of the image. The largest mean error, 0.24° and 0.08 mm, came with the least exposure and the incomplete image.

Matta and colleagues (International Journal of Implant Dentistry, 2026) built up gingiva over the scan bodies in two steps of about 3 mm and cut defects into them. Every error increased deviation: 30 µm in the best group with lower tissue, 178 µm in the worst with higher tissue.

Both studies used one- and three-implant models rather than full arches, but the lesson carries over: expose as much of the scan body as you can and leave no gaps in the mesh.

Which scan errors send a full-arch case back?

Each of these errors starts in the scan or on the order, and each one takes a quick check before the patient leaves.

How do you record the bite on an edentulous patient?

On an edentulous patient, record the full-arch bite with the provisional screwed in: it holds the vertical dimension and contacts you approved, and a buccal scan on both sides carries that relationship into the file. The same provisional, scanned out of the mouth, gives the technician the tooth form.

Without a fixed provisional, the reference has to come from somewhere else: the denture the patient wears, if you have approved its vertical dimension, or a record base with a wax rim. Agree with the lab on which one when you open the case. Immediate loading follows a different logic, with the scan taken right after surgery; our article on what the lab needs for immediate loading covers it.

Scan bodies or impression copings?

Scan bodies and impression copings both work for a full arch, and the choice is yours. In the systematic review by Gehrke and colleagues (International Journal of Implant Dentistry, 2024), scan bodies matched conventional impressions for single units and short spans. For full arches, there weren’t enough clinical data for a recommendation.

The meta-analysis by Pesce and colleagues (Journal of Clinical Medicine, 2025) pooled 9 clinical full-arch studies. The results conflict, but most put prostheses made from scans within clinically acceptable limits.

For scans, Gehrke’s review lists what changes accuracy: scan body design, material and color, scanner type, technique, scan aids, inter-implant distance, scan span and number of implants. The case brings some of these; technique and scan body choice are yours.

If the implant is deep and the scan body ends up covered, an impression with impression copings is the alternative. Send the copings or analogs with the order; we digitize the cast at the lab, and the design follows the same digital workflow.

Does photogrammetry replace intraoral scanning?

Photogrammetry doesn’t fully replace intraoral scanning. It’s an extraoral camera that records implant positions from dedicated scan bodies with optical targets, but it doesn’t capture gingiva, ridge or teeth, so the case still needs an intraoral scan or an impression.

In the Pozzi review (2025), covering 13 studies (3 clinical, 10 in vitro), photogrammetry was more accurate than intraoral scanning in 10 of the 13. The authors call for clinical trials before drawing conclusions. If you use photogrammetry, ask the lab before the first case how to send the camera files alongside the intraoral scan. The other reviews, the clinical studies and the chair time of each technique are in our article on photogrammetry vs intraoral scanning for full-arch implants.

When should you ask for a verification jig or bar try-in?

A try-in makes sense when the definitive is rigid and full-arch, like a milled titanium bar or a full-arch zirconia bridge. It’s what Pozzi and colleagues (2025) recommend: until clinical trials settle full-arch scan accuracy, try in a rigid prototype before the final prosthesis.

A verification jig is a rigid index that splints the implants in their recorded position; a bar try-in does the same job with the bar itself. In both, use the one-screw test (the Sheffield test): tighten only one terminal screw and look for rocking or gaps at the other abutments, then repeat from the other end.

Eid and colleagues (BMC Oral Health, 2026) used this test on a four-implant model, with the terminal screw at 15 Ncm, to measure the gap at the other three under titanium bars. Milled bars averaged 49 µm and printed bars 44 µm, both clinically acceptable. A try-in is there to find the gap while it can still be fixed; agree on it when you open the case.

The full try-in sequence, the other passive-fit tests and what to do when the bar won’t seat are in our framework try-in guide.

What does the lab check when the file arrives?

Smile Digital Lab, a digital dental lab in São Paulo, Brazil, making restorations for oral rehabilitation, checks at intake that the file opens and is complete, with every scan body readable, before the case moves on to design. If a scan body isn’t readable, the case stops there and we come back to you rather than design on a doubtful position.

We check every case when it arrives and on eight points before it ships, and you follow its phase in the portal (the screens are in Portuguese). We use Neodent and Titaniumfix components, both made in Brazil; if your system is different, put brand and platform on the order. The bridges themselves are on our zirconia on titanium bar page and full-arch zirconia page, and file basics are in our intraoral scanning guide.

Checklist before you send

• Provisional scanned in the mouth, with the full opposing arch and a bite on both sides
• Soft tissue scanned without scan bodies
• Every scan body complete, fully seated and clear of tissue at the base
• Provisional scanned out of the mouth, intaglio included
• Implant brand and platform, cuff height and angle of each abutment, scan body model
• Bar or prototype try-in agreed, when the definitive is rigid