What a Guide Lab Needs From Your CBCT and Intraoral Scan
Updated: Sep 24
Clinically reviewed by Jacob Shrayman, DDS, a general dentist who places implants using TrueGuide surgical guides in his own practice.
A guided case lives or dies on two files. Everything a laboratory does after that — the plan, the guide, the provisional — is built on a CBCT and an intraoral scan, and neither one gets better after the patient leaves the chair.
This is the record specification we work from at TrueGuide: what each file has to contain, the settings that decide whether it can be planned from, and the handful of things that send a case back. Which records a particular patient needs is a clinical decision and it stays yours. What follows is what the laboratory needs to build to your plan.
The CBCT
Send uncompressed DICOM, the whole multi-file series your unit exports, zipped. Not a viewer package, not a PDF report, not a compressed screenshot set, not a single flattened file. If your unit exports a viewer by default, go back and export the raw DICOM series instead; a viewer package carries a rendering, not the data the plan is measured in.
Field of view
The field has to include the surgical site and the anatomy that limits where an implant can go: the maxillary sinus and the nasal floor above, the inferior alveolar canal and the mental foramen below. For a single posterior site a small field is enough as long as it reaches those landmarks. For a full arch we need the whole arch, and on the mandible we need the canal where it enters and where it exits.
Use the smallest field that still captures them. That is the position the American Academy of Oral and Maxillofacial Radiology takes in its implant selection-criteria statement, and a larger field costs both dose and resolution. A field cropped so tight that the nerve or the sinus floor falls outside the volume looks fine on screen and cannot be planned from, because the structure that sets the trajectory is not in the data.
Voxel size
An isotropic voxel of 0.3 to 0.4 mm is enough to plan a guide from, and most units land in that range on their standard implant protocol. Anything under a millimeter works; 0.3 to 0.5 mm is where we see most cases arrive and that is fine. Going finer than 0.3 mm buys nothing measurable for a guide and costs dose and file size. Above about 0.5 mm the cortical outline starts to soften at the crest, which is where the plan needs it sharpest.
The ITI systematic review of CBCT imaging for implant planning reached the same conclusion: a voxel of 0.3 to 0.4 mm gives images of acceptable diagnostic quality, and smaller voxels did not consistently improve measurement accuracy at edentulous sites.
Scan geometry and reconstruction
Set the gantry tilt to zero. A tilted acquisition distorts the reconstruction and the volume cannot be used. Align the occlusal plane to the scan plane as closely as the chair allows.
Do not change reconstruction parameters partway through a series. An inconsistent slice increment is one of the most common reasons a data set will not import cleanly, and it fails on every planning platform, not only on ours. Export one series, at one setting.
If your software offers a metal artifact reduction filter, send the unfiltered series as well. The filter can soften the cortical outline the plan is measured against, and we would rather have both and choose.
Motion
Stabilize the head and tell the patient not to swallow during the acquisition. Motion shows up as a doubled or smeared cortical outline. If it falls anywhere near the surgical site the volume has to be retaken, because there is no way to know which of the two outlines is the bone. Look at the reconstruction before the patient leaves the operatory rather than after.
Arches open or closed
For a dentate case, have the patient bite on cotton rolls or a radiolucent bite stick so the arches are separated, and have them keep the lips and cheeks relaxed rather than pursed. Metal restorations throw streak artifact across the plane they sit in. When the arches are touching, the scatter off an opposing crown lands on the ridge you are trying to read, and there is no way to separate the two surfaces afterward. A few millimeters of separation costs nothing and keeps that scatter off the site.
The dual scan for an edentulous case is the opposite, and it is covered below: the patient wears the denture and closes into a radiolucent index rather than biting on cotton rolls.
Removable appliances
Take out partial dentures, removable retainers and any other removable appliance before a standard CBCT. Metal frameworks scatter, and acrylic gives a surface the software cannot separate from soft tissue. The only exception is the dual scan, where the denture is the point of the scan and stays in.
Where the Volume Starts and Stops
The height of the volume is what the plan is measured in, and it is the dimension most often clipped.
On the mandible, the volume should start below the inferior border and finish above the occlusal plane or the crest of the ridge, so the canal is captured where it enters at the mandibular foramen and where it exits at the mental foramen, along with any anterior loop.
On the maxilla, it should start below the occlusal plane and carry above the nasal floor and the sinus floor with a few millimeters to spare. A volume that clips the nasal floor cannot be used to plan an anterior maxillary case, no matter how good the resolution is.
When the Arch Is Full of Metal
If the arch carries multiple metal restorations, the scatter usually costs you the alignment between the CBCT and the intraoral scan rather than the bone reading itself. That is worth knowing, because the two problems have different answers.
The guide's fit surface comes from the intraoral scan, not from the CBCT. A clean scan of a heavily restored arch still gives us a seat. What the scatter threatens is the merge: the software aligns the two data sets on tooth surfaces, and if there is no clean tooth surface left to align on, the merge is where the case fails.
Where the arch is restored end to end, we can register the two data sets on radiopaque fiducial markers instead of on tooth surfaces. Tell us before the scan and we will send the marker protocol. If the scatter sits directly over the surgical site, we may still ask for a re-scan, and we will tell you that before you bring the patient back rather than after a plan has been built on it.
The Intraoral Scan
STL, PLY or OBJ, from any scanner. STL is the default; use PLY or OBJ if you want color carried over. We accept files from every intraoral system on the market, and a full-arch PVS impression or a poured model works too if that is how you practice — we digitize it here.
Scan the arch being treated, the opposing arch and the bite. Without the opposing arch and a bite we are planning an implant without knowing what goes on top of it. Scan the teeth the guide will sit on clean and dry; a wet or debris-covered seat surface is the one part of the scan a guide cannot forgive. If the plan has to work around an existing restoration or a provisional, scan that too.
Photographs
Send photographs on every full-arch case and on anything in the esthetic zone: retracted in occlusion, retracted with the teeth apart, an occlusal view of the arch being treated, and a natural smile.
They are the only record that carries lip position, the smile line and the existing vertical dimension. On a single posterior site they are optional. On anything the patient will see, they are what tells us where the tooth has to be before we decide where the implant can go. A full-arch plan built without them puts implants in acceptable bone and unacceptable teeth.
Existing Implants and Scan Bodies
If there are implants already in the arch that the plan has to reference, seat a scan body on each one and write the system and platform on the Rx.
Use the scan body made for that system and platform, hand-tighten it, and confirm it is fully seated before you scan. If you cannot see the seat, take a periapical. A scan body that is cocked or short of the platform puts the implant in the wrong place in the file, and the error is invisible once it is there. Without scan bodies, the position is being read off the CBCT, and that is not good enough to build to.
The scan body records position and rotation. Only the Rx tells us what connects to it.
Full-Arch and Edentulous Cases
An edentulous arch has no landmarks for the software to align to, so a full-arch case needs more than the two files. Which guide the case gets — tissue-supported and pinned, bone-supported, or a stackable base — changes what else we need, and all of them start here.
If the patient has a well-fitting denture the plan can be built around, use the dual-scan protocol. The full checklist:
The denture has to fit. Hard-reline it first if it rocks, because a loose denture produces a guide that does not seat. No metal, no mesh, no soft liner, no wax rim.
Place six to eight radiopaque markers around the arch, at least two per sextant, gutta-percha packed about a millimeter deep or the adhesive kind. Below the CEJ line, off the occlusal surface, off the borders and off the intaglio, and spread unevenly so the software cannot mirror them.
Take a radiolucent bite index, seat the denture on it, and scan the patient in occlusion — not open on cotton rolls. Then scan the denture alone on a foam block, the same way up, markers untouched, at the same settings as the first scan. Send both in DICOM.
Aligning the two volumes on those markers is what puts the planned tooth position into the software instead of leaving us to guess at it. Check both reconstructions before the patient leaves. Markers that moved between the two scans, a denture that moved, a soft liner, metal in the denture, a plastic base instead of foam, or teeth in contact on the first scan are what send a dual scan back.
If there is no denture worth building around, the case is planned from a pinned workflow instead, and we will tell you which before you take another record.
How Recent the Records Have to Be
The scan has to match the mouth on surgery day, not on the day it was taken. A temporary replaced, a tooth restored, a tooth extracted or tissue that has remodeled since the scan means the guide will match the record and still not seat, and that is the most common reason a guide does not fit.
Re-scan after any change to the teeth or the tissue the guide will sit on. A CBCT a few months old is fine if nothing in the mouth has changed; if there has been an extraction or a graft since, it is a new scan. We hold an unapproved case for thirty days, and if the scan is older than three months we will ask for a new one anyway.
The Rx
The Rx is the record that turns files into a plan, and it is the one most often short.
Tooth numbers. Implant system, platform, and the guided kit you own, by name. The diameter and length you have in mind. Pilot, drill-guided or fully guided. Flap or flapless. Immediate or delayed. Which teeth are being extracted, because a guide cannot take its support from a tooth that is coming out. What is going on top. And the surgery date, at the time you send the case rather than after the plan is approved.
The guided kit matters as much as the system. The sleeve, the vertical offset and every drill depth printed on the surgical report are kit-specific, so a guide built for a kit you do not own arrives with numbers you cannot use.
Check the Opening Before We Design
Measure the patient's maximum opening and tell us the site. A fully guided stack in the molar region wants roughly 50 mm of opening once you add the sleeve height, the guide thickness, the tissue, the drill above the sleeve and the handpiece head; a premolar site wants roughly 32 mm. The exact requirement is set by your implant system, not by us.
That is worth putting next to what patients actually open at the site — a mean of about 46 mm at incisor sites, 37 mm at premolar sites and 31 mm at molar sites. The gap is why guided molar cases fail in the chair more often than anywhere else in the arch.
If the room is not there, we would rather shorten the vertical offset, specify a shorter guided drill, move to an open-side or keyless sleeve, or plan that one site pilot-only with the rest fully guided, than have you find out with the guide in the mouth and the patient open.
Sending the Files
Every intraoral system on the market exports an open mesh. On iTero, scan as an iRecord rather than an aligner case and export the open STL. TRIOS exports through 3Shape Communicate, CEREC through Connect, Medit through Medit Link, Planmeca through Romexis. We also connect as your laboratory inside the scanner software, in which case cases route to us directly — tell us which scanner you run and we will send the connection details.
Zip the DICOM folder and label it with the patient's initials or your chart number rather than the full name; a surname in a filename survives every download, forward and backup of that file. Upload it with the meshes, the photographs and the Rx through the Submit a Case page. The upload is HIPAA compliant.
Do not email DICOM. It is too large for email and it is protected health information. We will sign your business associate agreement or send you ours, whichever your office prefers.
What Stops a Case
Five things account for nearly everything we send back, and all of them are fixable chairside, most at the same visit.
A field of view that misses the nerve or the sinus. A dentate CBCT taken with the arches in contact, so scatter off the opposing crowns lands on the ridge. A viewer package exported instead of the DICOM series. A scan body that was not fully seated, so the implant position in the file is not the implant position in the mouth. And an Rx that names the implant brand but not the platform or the guided kit, which is what sets the sleeve and the vertical offset, and with them every drill depth on the report.
A technician checks the merge between the CBCT and the scan before any planning begins. If the two data sets will not align, you get a call telling you exactly which record to retake, not a plan built on a bad alignment.
Who Reads the Volume
A CBCT for implant planning is justified because cross-sectional imaging orthogonal to the site is the only way to see ridge width and the distance to the canal or the sinus. That is the position the American Academy of Oral and Maxillofacial Radiology takes in its implant selection-criteria statement. The same statement asks for the smallest field of view that answers the clinical question, and it asks that the whole volume be systematically evaluated for findings, by the referring dentist or by an oral and maxillofacial radiologist.
Our technicians plan from the volume and flag anatomy near the surgical site. That is not a radiographic interpretation, and that duty does not transfer to the laboratory by sending us the file.
After You Upload
The case goes into planning, and the plan comes back to you for review before we print anything. You see the implant positions, the trajectory and the restoration the positions were planned from, and you can change any of it.
Revisions are unlimited and cost nothing until the guide prints. Once you approve the plan, a standard guide ships in 48 hours. Stackable and other multi-layer full-arch guides depend on the design and the material, and the schedule is set with you at the plan review.
Nothing prints without your approval. That is the whole point of reviewing a plan rather than receiving a guide.
Common Questions
Can you plan from a CBCT alone, with no intraoral scan?
No. The CBCT gives bone; the scan gives the soft tissue and the occlusal surfaces the guide sits on. A guide designed from bone data alone has nothing accurate to seat against.
What format does the CBCT have to be in?
Uncompressed DICOM, the whole folder of slices, zipped. Not a viewer package, not a single compressed file. If your unit exports a viewer by default, export the raw DICOM series instead.
What voxel size or slice thickness do you want?
Anything under a millimeter works. Most units land at 0.3 to 0.5 mm on their implant protocol and that is fine. Do not go down to 0.1 mm — it is a huge folder and it buys nothing for a guide.
What field of view do you need?
The area you are working in, plus whatever landmark constrains it. For a single site a small field is fine. For a full arch we need the whole arch, and on the mandible the canal where it enters and where it exits. A bigger field costs dose and resolution.
Bite open or closed for the scan?
For a dentate case, open — cotton rolls or a bite stick. If the teeth are touching we cannot separate the arches. For an edentulous dual scan it is the opposite: the denture in, the patient closed on a radiolucent index.
Can the patient wear a partial denture during the scan?
Not for a standard scan. Metal scatters, and an acrylic partial gives us a surface we cannot separate from tissue. Take it out. The exception is a dual scan, where the denture is the point.
What is the gantry tilt supposed to be?
Zero. A tilted scan distorts the reconstruction and we cannot plan from it.
The patient has a lot of crowns. Will the scatter kill the case?
Usually not. Scatter hurts the match between the CBCT and the intraoral scan more than it hurts the bone reading, and the guide's fit surface comes from the intraoral scan, not from the CBCT. Send it and we will tell you whether we can plan from it before you bring the patient back.
The patient moved during the scan. Do I retake it?
If the blur is at the surgical site, yes. We will not plan on a doubled outline, because there is no way to know which outline is the bone.
Which scan formats do you accept?
STL, PLY or OBJ. STL is the default; PLY or OBJ if you want color carried over.
Which scanners do you work with?
All of them. If it exports STL, PLY or OBJ we can plan from it, and we connect as your laboratory inside every major scanner platform if you would rather cases route to us directly.
I am on iTero. How do I get the file to you?
Scan it as an iRecord rather than an aligner case, then export the open STL. Or we connect as your laboratory in iTero and cases route straight to us.
TRIOS, Primescan, Medit, Planmeca?
All of them export an open STL — TRIOS through Communicate, CEREC through Connect, Medit through Medit Link, Planmeca through Romexis. Export and upload, or connect us as your laboratory.
Do you accept impressions instead of a scan?
Yes. Send a full-arch PVS impression or a poured model with the opposing arch and a bite, and we digitize it here.
Do you need the opposing arch?
Yes, and a bite. Without them we are planning the implant without knowing what goes on top of it.
How large are these files, and how do I send them?
A CBCT folder is typically a few hundred megabytes. Zip it, label it with the patient's initials or your chart number, and upload it with the meshes and the Rx on the Submit a Case page.
Can I email the DICOM?
No. It is too large for email and it is protected health information. Use the upload on the site.
Is the upload HIPAA compliant?
Yes. And we will sign your business associate agreement or send you ours, whichever your office prefers.
The patient already has implants in the arch. What do you need?
A scan body on every implant the plan has to reference, and the system and platform on the Rx. Without scan bodies their position is being read off the CBCT, and that is not accurate enough to build to.
How do I know a scan body is seated?
Hand-tighten it, look at the seat, and if you cannot see it, take a periapical. A scan body that is cocked or short of the platform puts the implant in the wrong place in the file, and the error is invisible once it is there.
What is a dual scan, exactly?
Two CBCTs. First the patient wearing a metal-free denture with six to eight radiopaque markers, closed on a radiolucent index. Second the denture alone on a foam block, same orientation. We align the two volumes on the markers.
Where do the dual-scan markers go?
Six to eight around the arch, at least two per sextant, below the CEJ line, spread unevenly, off the occlusal surface, off the borders and off the intaglio. They must not move between the two scans.
Why did our dual scan get sent back?
The usual reasons: markers moved between the scans, the denture moved, it was scanned on plastic instead of foam, there was a soft liner or metal in it, or the teeth were in contact on the first scan.
The denture rocks. Can we still use it?
Hard-reline it first. A loose denture produces a guide that does not seat, and no amount of planning fixes that afterward.
Do you need photographs?
On every full-arch case and anything in the esthetic zone: retracted in occlusion, retracted with the teeth apart, an occlusal view, and a natural smile. On a single posterior site they are optional.
How recent does the scan need to be?
It has to match the mouth on surgery day. If a tooth was restored, extracted or moved after the scan, the guide will not seat. Re-scan after any change to the teeth or tissue the guide sits on.
How long do you hold a case I have not approved?
Thirty days. After that it is archived and we would start it fresh, and if the scan is older than three months we will want a new one anyway.
What has to be on the Rx?
Tooth numbers, implant system, platform, guided kit, planned diameter and length, pilot or drill-guided or fully guided, flap or flapless, immediate or delayed, which teeth are being extracted, what goes on top, and the surgery date.
Why do you need to know which guided kit I own?
Because the sleeve, the vertical offset and every drill depth on the surgical report are kit-specific. A guide built for a kit you do not own comes with depths you cannot use.
Do I need to measure the patient's mouth opening?
For a posterior fully guided site, yes. A fully guided stack in the molar region wants roughly 50 mm and a premolar site roughly 32 mm, and the mean opening available at a molar site is about 31 mm. Measure it and send it with the case.
Who reads the CBCT?
Our technicians plan from it and flag anatomy near the surgical site. That is not a radiographic interpretation. The whole volume still has to be evaluated for findings by you or by an oral and maxillofacial radiologist, and that responsibility stays with the treating doctor.
What happens after I upload?
A technician checks the merge between the CBCT and the scan, the case goes into planning, and the plan comes back to you before anything prints. Revisions are unlimited and free until it prints, and a standard guide ships 48 hours after you approve.
Send a Case
Upload your CBCT, intraoral scan, bite, photographs and Rx on the Submit a Case page, or call (201) 255-4233 and we will tell you which records the case needs before you take another one.
References
Bornstein MM, Horner K, Jacobs R. Use of cone beam computed tomography in implant dentistry: current concepts, indications and limitations for clinical practice and research. Periodontol 2000. 2017;73(1):51-72.
Fokas G, Vaughn VM, Scarfe WC, Bornstein MM. Accuracy of linear measurements on CBCT images related to presurgical implant treatment planning: a systematic review. Clin Oral Implants Res. 2018;29(Suppl 16):393-415.
Tyndall DA, Price JB, Tetradis S, et al. Position statement of the American Academy of Oral and Maxillofacial Radiology on selection criteria for the use of radiology in dental implantology. Oral Surg Oral Med Oral Pathol Oral Radiol. 2012;113(6):817-826.
Gargallo-Albiol J, Ortega-Martinez J, Salomo-Coll O, et al. Vertical mouth opening and fully guided implant surgery. Int J Oral Maxillofac Surg. 2024;53(6):526-532.
Tahmaseb A, Wu V, Wismeijer D, Coucke W, Evans C. The accuracy of static computer-aided implant surgery: a systematic review and meta-analysis. Clin Oral Implants Res. 2018;29(Suppl 16):416-435.
Comments