03 — Scanning
The scanner measures.
It does not photograph.
A clinical and scientific guide to turning digitisation into reliable data: intraoral scanner, photogrammetry, benchtop scanner, crowns, occlusal devices, surgical guides, edentulous areas, removable prostheses, implants and digital case visualisation.
The scanner is not an ordinary camera
Scanning well means controlling geometry, light, tissue, saliva, movement and strategy.
The intraoral scanner captures a sequence of images, points or optical frames and reconstructs a three-dimensional mesh. When the operator loses anatomical reference, moves the scanner without a sequence, or scans saliva, blood, mobile tissue or an uncontrolled reflective surface, the software can stitch the data together unstably.
Underlying the process is a geometric logic: the system estimates three-dimensional coordinates, organises that data as a point cloud and, from it, reconstructs a polygonal mesh that will be exported as STL, PLY, OBJ or a proprietary format. In simple terms, the scanner does not photograph the mouth; it measures surfaces and turns those measurements into a three-dimensional model.
Clinical technique therefore remains decisive. Premium equipment does not compensate for a poorly exposed preparation, a wet subgingival margin, a badly seated scan body, inadequate retraction, or the absence of reference points in extensive edentulous regions.
Scanning technologies
Commercial intraoral scanners use distinct optical principles to estimate depth and reconstruct the surface. The final clinical experience depends on the combination of hardware, reconstruction algorithm, field of view, ergonomics, software, computer, scanning strategy and system updates.
Triangulation is a geometric principle: if the system knows the relative position of the light source and the sensors, the deformation or displacement of the reflected pattern allows the spatial position of the surface to be estimated. The initial result can be understood as thousands or millions of three-dimensional coordinates — a point cloud. The software then filters noise, aligns successive frames, fills small areas and converts the points into a triangular mesh. That mesh is what normally reaches the CAD software, the laboratory or the 3D printer.
Not every manufacturer publishes the full optical architecture of its equipment in detail. The examples below are therefore described conservatively, using public manufacturer information and the scientific literature.
Triangulation / structured light
Projects light or luminous patterns onto the surface and computes depth from the geometry between source, camera and object. Capture generates spatial coordinates that form a point cloud before becoming a mesh. It is sensitive to reflection, moisture and lack of texture, but can be fast and efficient when the field is clean.
Confocal imaging
Reconstructs the surface from focal planes. It generally allows powder-free capture in many modern systems and can work well on complex anatomies, provided the field is dry and the strategy consistent.
Active wavefront sampling
Estimates depth by analysing variations of the optical beam and focus during capture. It is one of the approaches described in the literature for three-dimensional reconstruction in digital dentistry.
Video, tracking and stitching
Many scanners capture data continuously. The software identifies repeated structures, registers successive frames, reduces noise and creates the final mesh. The larger the arch and the fewer the anatomical references, the greater the risk of accumulated error.
Colour, texture, NIRI and diagnostic features
Some systems add real colour, texture, change maps over time, near-infrared transillumination or follow-up features. This data helps with communication and documentation, but does not replace clinical and radiographic diagnosis.
Photogrammetry as a complementary resource
In full-arch implant-supported rehabilitations, photogrammetry systems can record the spatial position of implants with high consistency. It is not the same as scanning teeth and tissue; it is a complementary workflow for specific cases.
| Technology | Documented examples | Clinical note |
|---|---|---|
| Triangulation — structured light or video | Medit i500; the CEREC Omnicam is described in the literature as a video system with active triangulation. | Control of brightness, moisture and wand angle is decisive. Metallic or highly polished surfaces may require a change of angulation and selective recapture. |
| Confocal / optical sectioning | iTero Element/Element 5D; 3Shape TRIOS, described by the manufacturer as Ultrafast Optical Sectioning™ technology. | Can be very efficient on preparations and areas with rich anatomy, but remains dependent on an exposed margin, a dry field and a stable path. |
| Active wavefront sampling | 3M True Definition Scanner, described in the literature as based on pulsed blue light and active wavefront sampling. | An important historical example for understanding the evolution of optical systems, even though some devices have been discontinued. |
| Dynamic depth scan | Dentsply Sirona Primescan; the manufacturer describes dynamic depth scan technology and processing of more than one million 3D points per second. | The proposal is to ease deep areas and difficult materials, but field quality and a final check remain indispensable. |
| NIRI / auxiliary imaging | iTero Element 5D Plus uses NIRI; Primescan 2 also advertises 2D fluorescence and near-infrared imaging, subject to regional availability. | These features help with communication and documentation. They do not turn the scanner into a stand-alone substitute for clinical, radiographic or formal diagnosis. |
| Photogrammetry | PIC dental, ICam/Imetric and SHINING 3D IPG are examples of systems aimed at photogrammetry and implant position. | Not the same as scanning teeth and mucosa. In implant-supported full-arch cases it can complement the intraoral scanner to reduce implant position error. |
Photogrammetry: phone, camera and implants
Three-dimensional reconstruction from multiple two-dimensional images taken from different positions.
The software identifies homologous points between photographs, estimates the relative camera position, generates a sparse point cloud, densifies it and creates a textured mesh. The idea resembles human vision: two or more perspectives allow depth to be inferred.
In dentistry the term appears in different contexts. There is dedicated dental photogrammetry, used mainly to record implant positions with specific scan bodies; and there is general photogrammetry, done with a phone, camera or apps, useful for plaster models, objects, facial documentation, teaching and communication. These two applications should not be treated as equivalent.
Plaster models and objects
A phone or camera can digitise plaster models, prototypes, printed parts, wax-ups, guides, trays and teaching objects. The application is safer for documentation, preliminary planning, digital archiving and teaching. For fine marginal fit, dies, abutments, implants or final prosthetic work, accuracy must be validated before replacing a benchtop scanner.
Face scanning
Smartphone apps, depth sensors and photographic sequences can generate 3D facial models for communication, aesthetic analysis, digital smile design, orthodontics, surgery and integration with intraoral scanning. The main value is contextual: the relationship between teeth, lips and face, symmetry, profile, smile and communication with the patient.
Scan bodies and full-arch
Dedicated systems such as PIC dental, ICam/Imetric and SHINING 3D IPG use specific scan bodies and their own algorithms to record the spatial position of implants. The objective is not to copy teeth or mucosa, but to determine implant coordinates with greater stability in extensive rehabilitations. Phone photogrammetry should not be confused with these systems.
Intraoral photos ≠ intraoral scanner
Intraoral reconstructions from photographs are promising in research and telemonitoring, but the mouth is a difficult environment: reflective enamel, saliva, shadows, tongue, limited access, few angles and no reliable scale. For prostheses, guides, occlusal devices and working models, the intraoral or benchtop scanner remains the more predictable workflow.
Basic protocol for photogrammetry of a model, object or face
- Use diffuse lighting, without hard shadows, intense glare or direct reflections.
- Take many images with high overlap between them; short, progressive movements work better than a few distant photographs.
- Keep the object still; if using a turntable, keep the camera fixed and the background rich in references.
- Avoid transparent, shiny, mirrored or overly smooth surfaces. On non-clinical objects, a matting spray can help; on intraoral items and patients, use only safe and indicated materials.
- Insert a scale, ruler, markers or dimensional reference when the use requires real measurement.
- Inspect for holes, deformed areas, excessive smoothing and distortion before using the mesh.
Low-cost photogrammetry is useful, but it should not be sold as a universal replacement for an intraoral scanner, a benchtop scanner or validated implant photogrammetry. The level of requirement changes with the use: documenting a face is different from manufacturing a passive framework over multiple implants.
Benchtop scanners: laboratory and models
Extraoral equipment used mainly in the laboratory or clinic to digitise plaster models, impressions, dies, occlusal records, wax-ups, provisionals, existing prostheses, scan bodies and prosthetic components.
Unlike the intraoral scanner, it works outside the mouth, in a more controlled environment, with a still object, controlled lighting and mechanical support. Benchtop scanners are therefore usually very predictable for models, dies and laboratory work. However, they do not eliminate errors accumulated before digitisation: a distorted impression, expanded plaster, a bubble, a tear, a worn model or a badly screwed scan body remain sources of error.
Structured light
Projects light patterns onto the object and reconstructs the surface from the deformation of those patterns. It is the predominant architecture in modern laboratory scanners, combining speed, automation and good ability to capture detail on models and dies. Commercial examples include the 3Shape E, Medit T, Shining 3D AutoScan and DOF Freedom series, varying by market and version.
Laser
Uses a laser beam or line to scan the object. It can be precise, but in many dental workflows it has been replaced by structured light for speed and practicality. It still appears in some general 3D scanners or older systems.
Axes, turntable and articulator
Models with more axes, a rotary table, automatic tilting and articulator accessories can capture retentive areas, dies and occlusal relationships with less manual intervention. The difference is not only resolution; it is the ability to expose angles the camera needs to see.
Scan bodies and libraries
In implant dentistry, the benchtop scanner can digitise models with analogues and laboratory scan bodies. Accuracy depends on component seating, torque, the correct CAD library, model quality and scanner strategy. For implant-supported full-arch cases, dedicated photogrammetry can be added when implant spatial position is the priority.
| Use | Strong application | Limitation | Practical note |
|---|---|---|---|
| Plaster model | Digital archive, laboratory CAD, fixed prosthesis, RPD, occlusal device and comparison with intraoral scanning. | The digitisation will only be as good as the physical model. Bubbles, expansion and wear are not corrected by the scanner. | A robust workflow when a good impression or model already exists, or when the laboratory needs to preserve conventional steps. |
| Impression | Some scanners accept direct scanning of impressions, reducing the need to pour plaster in certain workflows. | Material, colour, translucency, depth, bubbles and distortions of the impression influence the reading. | It can be useful, but must not mask a clinically deficient impression. |
| Removable prosthesis | Digitisation of an existing prosthesis, wax-up, functional model, custom tray and records for reverse planning. | It does not record mucosal compressibility or functional extension if that information is not physically present on the scanned object. | It is often more predictable to scan an existing prosthesis or a functional impression than to try to capture the whole functional periphery directly in the mouth. |
| Die and preparation | Well-exposed margins on dies and working models can be captured with high control of lighting and stability. | If the margin was not captured in the impression, or if the model has a bubble or tear, the scanner cannot recover missing information. | It remains a solid laboratory workflow for cases where the conventional impression was well executed. |
The intraoral scanner captures anatomy directly on the patient. The benchtop scanner captures physical objects outside the mouth with greater control, but depends on the quality of the impression or model. Dedicated photogrammetry records the spatial relationships of targets or scan bodies from multiple images, and is particularly relevant in extensive implant cases. They are complementary tools, not perfect substitutes for one another.
Accuracy, trueness and precision
Accuracy is not a synonym for a pretty image. In metrology, accuracy involves two components: trueness and precision.
In dental practice, this means the file must be close to the real anatomy and also repeatable when the same case is scanned again under similar conditions.
Accuracy
Overall closeness between the scan and clinical reality. It depends on the sum of systematic and random error.
Trueness
Closeness between the mean captured value and an accepted reference value.
Precision
Repeatability. A scanner can be precise, repeating the same error, and still not be true.
Resolution, mesh density, colour texture and the visual fluidity of the software do not guarantee clinical fit. A file can look excellent on screen and still have accumulated distortion, an incomplete margin, soft tissue captured as if it were rigid structure, or an unstable interocclusal relationship.
Flowchart of a good scan
Five steps that determine whether the file will be usable.
Prepare the field
Remove excess saliva, control bleeding, retract lips, cheeks and tongue. The scanner should read tooth and stable tissue, not fluid.
Create a reference
Start from areas with recognisable anatomy: cusps, fossae, ridges, adjacent teeth and non-reflective surfaces.
Follow a path
Use a continuous sequence. On dentate arches, a frequent pattern is occlusal, lingual or palatal and buccal, avoiding long jumps.
Validate the mesh
Inspect margins, contact areas, edentulous regions, mucosa and occlusal points. Erase and recapture whatever is poor.
Export correctly
Choose STL, PLY, OBJ or a proprietary file according to the objective: CAD, laboratory, communication, temporal comparison or planning.
Protocols by clinical indication
What must be in the file, case by case.
7.1Fixed prosthesis — crown, onlay, endocrown and veneer
- Fully expose the finish line before scanning.
- Control blood, crevicular fluid and saliva; a wet margin becomes an imprecise margin.
- Scan adjacent teeth and the antagonist with quality sufficient for proximal contact and occlusion.
- Review preparation areas, internal angles, cervical finish line and proximal surfaces before sending to the laboratory.
7.2Occlusion — occlusal device
- Capture complete arches, without holes on incisal edges, cusps and distal regions.
- Record the bite bilaterally where indicated, avoiding a single unstable buccal record.
- Check that the software has assembled the arches coherently before designing the device.
- Prefer files with good occlusal anatomy; excessive smoothing impairs adjustment.
7.3Implant dentistry — surgical guides
- The scan must have well-defined support areas: teeth, stable mucosa or auxiliary devices.
- Planning combines the optical surface with DICOM from the tomography; an error in either source compromises the guide.
- In extensive edentulous cases, assess the need for markers, a radiographic guide, a conventional impression or a hybrid protocol.
- Check the virtual and physical seating of the guide before surgery.
7.4Partial and complete edentulous areas
- The larger the smooth, featureless area, the greater the tracking challenge.
- Use remaining teeth, rugae, papillae, the palatal raphe and other anatomical landmarks where present.
- Avoid scanning mobile mucosa, pooled saliva and the vestibular fold without control.
- In fully edentulous arches, conventional or hybrid protocols may be more predictable depending on the objective.
7.5Removable prosthesis — RPD and complete denture
- For RPDs, capture rests, rest seats, guide planes, prosthetic equators and ridge areas.
- For complete dentures, the functional surface and the peripheral border remain a challenge for direct capture.
- A useful strategy is to digitise an existing prosthesis, a custom tray or a functional impression when the case requires a more faithful peripheral extension.
- Do not confuse a good-looking digital model with a functional record of the mucosa.
7.6Communication — visualisation and follow-up
- PLY and OBJ can preserve colour and texture, useful for communication, documentation and monitoring.
- Serial comparisons can reveal wear, recession, alignment changes and anatomical change.
- Clinical value depends on standardisation: the same scanner, the same protocol and critical interpretation.
- Scanning complements, but does not replace, clinical, radiographic and periodontal examination when needed.
Retraction and field control
Retractors do not merely create space. They stabilise lip, cheek and tongue, reduce saliva contamination and improve the scanner's line of sight. The choice depends on the arch, the type of procedure, patient tolerance and the area to be captured.
- Wide lip retractors: useful for full arches, orthodontics and general visualisation.
- Cheek retractors: help with posterior buccal capture and bite registration.
- Clinical mirror or tongue retractor: useful for tongue control, but must not obstruct the optical path.
- Suction, gauze and cotton rolls: moisture and field control, especially in posterior regions.
- Retraction cord, haemostatic paste or an equivalent technique: indicated when the preparation margin must be exposed.
Pooled saliva, bubbles, blood, moving soft tissue, a retractor touching the margin, retraction cord covering the cervical finish line, and a highly reflective metal surface without control can all generate artefacts. The right approach is to erase the defective area and recapture only once the field is under control.
Export formats
The file format is not a bureaucratic detail. It defines what will be preserved: geometry alone, geometry with colour, texture, relationships between meshes, or the software's proprietary data.
| Format | What it preserves | Advantages | Limitations | Typical clinical use |
|---|---|---|---|---|
| STL Standard Tessellation Language |
Surface geometry as a triangular mesh. | High compatibility with CAD, 3D printing and milling. It is the most widely accepted format in open workflows. | Preserves neither real colour nor texture. May be insufficient for fine visual communication. | Crowns, models, occlusal devices, guides and general CAD/CAM where colour is not needed. |
| PLY | Geometry and, when exported by the scanner, per-vertex colour information. | Useful for documentation, communication, monitoring and software that uses colour and texture. | Not every laboratory or CAD package handles PLY well. Compatibility must be confirmed. | Case visualisation, temporal comparison, digital orthodontics, communication with patient and laboratory. |
| OBJ | Geometry and texture through associated files, depending on the export. | Can retain richer texture when the workflow accepts the accompanying files. | Usually involves more than one file. Texture can be lost if the associated files are sent incomplete. | Visual communication, aesthetic planning and workflows that require texture. |
| DICOM | Volumetric radiographic data, not the intraoral optical surface. | Essential for implant planning, relationship with bone, mandibular canal, maxillary sinus and anatomical structures. | Does not replace STL, PLY or OBJ for surface anatomy. It must be aligned with the optical scan. | Surgical guides, implant dentistry, endodontics, surgery and diagnostic imaging. |
| Proprietary | May include the case, bite, stages, metadata, colour, history and integration with specific software. | A more complete workflow within the platform itself. | Lower interoperability. May depend on licence, account, version or subscription. | Communication within the scanner's ecosystem, a partner laboratory or platform-guided planning. |
Price tiers and a rational choice
Intraoral scanners should not be compared on purchase price alone. The real cost involves the computer, support, warranty, training, software licensing, open export, subscription, maintenance, tips, sterilisation, integration with the laboratory and clinical time saved.
| Tier | General profile | What to assess | Common risk |
|---|---|---|---|
| Entry | More affordable equipment, frequently sold as a lower initial investment alternative. | Open export, software stability, local support, speed, ergonomics, the computer required and full-arch performance. | Saving on the purchase and losing predictability through weak support, limited software or a poorly guided learning curve. |
| Mid-range | Systems with a better balance of cost, software, user community and integration with CAD/CAM workflows. | Licences, updates, cloud quality, libraries, laboratory integration and ease of exporting STL, PLY and OBJ. | Assuming that every mid-range scanner solves every case, including fully edentulous arches and implant-supported full-arch work. |
| Premium | More robust ecosystems, with mature software, diagnostic features, orthodontic or prosthetic integration and structured support. | Total cost, subscription, ecosystem lock-in, laboratory compatibility, cart or laptop, and maintenance cost. | Paying for features that will not be used, or being locked into a closed workflow incompatible with your routine. |
The tiers are conceptual. Values change with importation, exchange rate, taxes, distributor, warranty, software package and training. For a purchase decision, request a clinical demonstration and test it with the cases you actually perform.
Quality checklist before sending
Four checks by case type, before the file leaves the practice.
- Crown and indirect restoration
- Margin visible through 360°, no blood or saliva at the finish line, proximal contacts preserved, antagonist intact, coherent occlusal record and no artificial smooth areas over the margin.
- Occlusal device
- Complete arches, occlusal anatomy without gaps, a stable bite record, canines and incisors well captured, posterior areas without holes and no duplicated mesh.
- Surgical guide
- A well-defined support surface, an STL compatible with the DICOM data, sufficient reference points, no mucosal movement, and confirmation of the planned seating.
- Edentulous region
- A short, progressive path, use of anatomical landmarks, control of mobile mucosa, no thick saliva film, and critical validation before designing a prosthesis or guide.
Essential scientific basis
Manufacturer technical sources are identified as such and are not presented as scientific evidence.
- 3Shape. Dental laboratory scanners. Technical text on laboratory scanners, noting that some systems use laser, while structured light predominates in modern benchtop scanners. Technical source
- 3Shape. TRIOS — Ultrafast Optical Sectioning™. Technical distribution material. Kept as a technical source, not as a scientific article. Distribution technical material. Technical material (PDF)
- Al-Rudainy D et al. Smartphone-based photogrammetric technology for orthodontic dental casts. Study on the validity and reliability of 3D models of orthodontic casts obtained by smartphone photogrammetry. PubMed
- Borbola D et al. In vitro comparison of five desktop scanners and an industrial scanner. Study comparing desktop scanners and an intraoral scanner on complete-arch dentate models. PubMed
- Dentsply Sirona. Primescan. Official manufacturer page with clinical characteristics; for technical documentation, consult the brand's official materials as well. Official source
- Gehrke P et al. A systematic review of factors impacting intraoral scanning accuracy in implant dentistry. Systematic review of factors influencing accuracy in implant dentistry. Full text · PubMed
- ISO 5725-1:2023. Accuracy (trueness and precision) of measurement methods and results. Metrological reference for accuracy, trueness and precision. ISO
- Jindanil T et al. Smartphone applications for facial scanning. Technical review of smartphone applications for facial scanning, portability and limitations. Full text · PubMed
- Kihara H et al. Accuracy and practicality of intraoral scanner in dentistry. Review of accuracy, practicality and verification methods for intraoral scanners. PubMed
- Medit. i500 — video-type scanning and triangulation. Manufacturer's technical text explaining video-based triangulation. Manufacturer technical text. Technical source
- Mello LS et al. Smartphone photogrammetry: affordable alternative for digitizing dental casts. Low-cost protocol for digitising dental casts by smartphone photogrammetry. Page · PDF
- PIC dental / Imetric / SHINING 3D. Photogrammetry for implant registration. Public sources on photogrammetry for implant registration and complete-arch workflows. PIC · ICam · IPG
- Richert R et al. Intraoral Scanner Technologies: A Review. Classic review of technological principles, scanning path and mesh quality. Full text · PubMed
- Sacher M et al. Accuracy of commercial intraoral scanners. Open-access article describing and comparing commercial scanners, including active wavefront sampling. Full text · PubMed
- Srivastava G et al. Accuracy of intraoral scanner for recording completely edentulous arches. Review of scanning edentulous arches and the limits of poorly trackable structures. PubMed
- Vitai V et al. Evaluation of the accuracy of intraoral scanners for complete-arch scanning. Systematic review and network meta-analysis of complete-arch accuracy. PubMed