Journal of Orthodontics, Vol. 31, 2004, 154–162
`
`FEATURES
`SECTION
`
`Current Products and Practices
`
`Applications of 3D imaging in
`orthodontics: Part II
`
`M. Y. Hajeer
`University of Albaath Dental School, Damascus, Syria
`D. T. Millett and A. F Ayoub
`University of Glasgow Dental School, Glasgow, UK
`J. P. Siebert
`Department of Computing Science, Glasgow University, Glasgow, UK
`
`Introduction
`
`3D CT scanning
`
`The first part of this article has illustrated some of the
`principles in 3D imaging and the possible applications
`of 3D facial imaging in orthodontics. Part II explores
`the different techniques of 3D imaging of the teeth, as
`well as the recent efforts to create the ‘virtual orthodontic
`patient’ by using 3D soft and hard tissue data. A brief
`overview of some of the commercially available 3D-based
`technologies, such as OrthoCAD™ and Invisalign® is
`given at the end.
`
`3D imaging of the teeth
`
`3D laser scanning
`
`Intra-oral laser scanning may be difficult due to the pos-
`sibility of patient movement during scanning, in addition
`to the safety issues related to the laser.1 On the other
`hand, laser scanning of study casts has many advantages
`over other scanning techniques, despite the long time of
`acquisition. Obviously, time of exposure is not an issue in
`this type if imaging. The problem of capturing the mor-
`phology of a study cast is related to the presence of many
`areas of undercut, not to its texture. This can be solved by
`capturing the study model from several different angles,
`which enables the production of a 360° model with a very
`high accuracy (Figure 1). Once the 3D model has been
`produced, the operator can save it in the hard disc of a
`computer in a specific 3D file format and the size of this
`file is dependent on the original resolution of the 3D
`mesh. There are different 3D file formats such as VRML
`(.wrl), which stands for Virtual Reality Modelling
`Language, and .stl for Stereolithographic formats, and
`.dxf, which is one of the formats used by the AutoCAD
`program.
`
`Address for correspondence: Dr Mohammad Y Hajeer, Orthodontic
`Department, University of Albaath Dental School, PO Box 9309, Mazzeh,
`Damascus, Syria. Email: hajeer@scs-net.org
`© 2004 British Orthodontic Society
`
`3D CT scanning is another option, but its cost limits its
`usefulness in daily clinical practice. A validation of the
`process is required to estimate the error, since study cast
`stone is more radiodense than bone.
`
`Stereophotogrammetry
`
`Although this technique has proved to be very valuable in
`imaging human faces,2 it is not so suitable for capturing
`study casts. Ayoub et al.3 discussed the possibility of
`employing this technique to archive study casts in orth-
`odontic practice and proposed a specific configuration
`of the system to achieve high quality models with an
`estimated accuracy of 0.2 mm.
`
`Intra-oral direct dental scanning
`
`OraScanner™ the first 3D hand-held intra-oral scanner,
`has been developed by OraMetrix Company in the USA,
`and depends on the structured light technique. A video
`camera records the structured light distortions on the
`dental crowns as it passes over the dentition in about one
`minute.1 The computer processes these images and
`merges them together to create a complete 3D dental
`arch.
`
`Applications of 3D imaging of the
`teeth
`• Archiving ‘study casts’. 3D images are a reliable way
`to archive study models, producing durable images
`without any fear of loss or damage to the original casts.
`If a model requires 5 Mb of space, one CD-ROM can
`accommodate between 130 and 145 study casts. A hard
`
`DOI 10.1179/146531204225020472
`
`Align EX1018
`Align v. 3Shape
`IPR2022-00145
`
`

`

`JO June 2004
`
`Features Section
`
`3D imaging in orthodontics
`
`155
`
`Figure 1 A 3D study model captured by laser scanning. The density of the meshes should be high enough to reveal accurate occlusal morphology.
`Time of data acquisition is long, but the accuracy of the final output is superior to other techniques.
`
`disc of a 60-Gb capacity can accommodate approxi-
`mately 12,000 study models. It has been calculated that
`for a British unit, which sees 1000 new patients each
`year, 17 m3 of storage space are required for study
`model storage.4 With the use of 3D digital study
`models, such storage space is no longer needed.
`• Documentation of treatment progress and communi-
`cation between professional colleagues is also made
`easier by examining records in 3D. Tele-orthodontics
`saves time and effort in treatment planning, since the
`need for a physical meeting is no longer required.
`• With new advances in 3D dental and orthodontic
`software, the orthodontist can examine intra- and
`inter-arch relationships with much more precision.
`Transverse relationships between upper and lower
`arches could be better evaluated when 3D models are
`viewed in occlusion from different angles on the screen.
`Treatment objectives and treatment planning can be
`created taking into account the different treatment
`options, ending with what could be termed ‘virtual
`
`treatment’ and ‘virtual set-up’ of the orthodontic
`appliance. More details are mentioned below about
`OrthoCAD™ technology.
`• Simulation of space closure following extraction, tooth
`uprighting or incisor retraction can be easily shown
`to patients, which increases their understanding and,
`perhaps, their compliance.
`• 3D prefabrication of archwires using specific robotics
`after setting up bracket positions on the dental arches.
`The reader is referred to the OraMetrix website.5
`• Construction of 3D ‘aligners’, which are thin, clear,
`overlay appliances used in a sequential manner over a
`period of time to correct a malocclusion without the
`need for conventional fixed appliances. The Invisalign®
`technology is discussed later.
`
`Virtual orthodontic patient
`
`The ultimate dream of 3D imaging and modeling is to
`achieve the ‘virtual orthodontic patient’, where we can
`
`

`

`156 M. Y. Hajeer et al.
`
`Features Section
`
`JO June 2004
`
`see the bone, flesh and teeth in three dimensions. If this
`can be achieved in an accurate way, it will allow consider-
`able data to be collected and a variety of soft and hard
`tissue analyses to be performed. Our knowledge of the
`masticatory system will increase, and our understanding
`of tooth movement biomechanics, orthopedic and
`orthognathic corrections will be enhanced.
`From the historical point of view, Calvin Case6 was the
`first to criticize the Angle classification because he
`thought it was an over-simplification of the diverse varia-
`tion in types of malocclusion.1 Van Loon, in 1915, agreed
`that a 3D system was required to determine the relation-
`ship of the dentition to the face for meaningful diagnosis
`and treatment planning.7 He used a plaster cast of the
`face along with a plaster cast of the dentition. Simon, in
`1922, tried to relate the study models to the craniofacial
`complex. His apparatus included a maxillary clutch and
`frame, which was similar in its principle to the later
`facebows of Hanau and McCollum.8 Both van Loon
`and Simon’s contributions were very important because
`they were among the first to stress the need for under-
`standing the spatial relationships between the different
`components of the craniofacial complex.9
`Despite the drawbacks of CT scanning, attempts have
`been made to combine 3D skeletal hard-tissue informa-
`tion derived from CT scans with other 3D dental and/or
`soft-tissue information obtained by vision- or laser-based
`scanning techniques. Recently, Xia et al.10 developed a
`system for reconstructing 3D soft and hard tissue models
`from sequential CT slices using a surface rendering
`technique followed by extraction of facial features from
`3D soft tissues. Three digitized color portraits were
`texture-mapped onto the 3D head mesh. Although this
`technique was interesting in showing the importance of
`having the full color details of patients’ faces in the final
`output, the validity of the construction process was not
`evaluated, nor the potential error of facial expression
`change during data acquisition.2
`A combination of 3D CT skeletal maps with 3D
`laser-based study models was attempted by Nishii et al.11
`and Terai et al.,12 but they reported significant errors
`in positioning. Other researchers tried to combine 3D
`skeletal data based on cephalograms with 3D laser-
`scanned dental models in order to overcome the problems
`associated with CT skeletal data.13 However, this tech-
`nique cannot be used for prediction of soft tissue changes
`following treatment, which minimizes its applicability.
`On the other hand, Chen and Chen14 employed 3D ceph-
`alometric skeletal data in conjunction with 3D facial
`soft-tissue data derived from laser scanning to achieve
`a 3D computer-aided simulation system to plan surgical
`procedures and predict postoperative changes
`in
`orthognathic surgery patients.
`
`A major focus of the Craniofacial Research Instrumen-
`tation Laboratory (CRIL) at the University of the Pacific
`School of Dentistry has been the development of instru-
`mentation, software and procedures for the creation of
`an accurate, integrated 3D craniofacial data model.15
`To merge different 3D maps representing different
`craniofacial structures, they have emphasized the need
`for ‘tie points’, which are landmarks placed on specific
`areas on the face prior to imaging.8 The anatomic
`features located on a stereo X-ray image act as a frame-
`work on which data from other sources (3D study cast
`models, 3D facial images) are hung. 3D facial models
`are acquired using a structured light technique, while
`3D study models are built using ‘destructive scanning’
`machines. No validation of the whole construction and
`integration procedures has been reported yet.
`A new method of combining and mapping patients’
`facial textures (based on stereophotogrammetry) onto
`3D spiral CT skeletal and soft-tissue data was proposed
`recently by Khambay et al.16 (Figure 2). However, this
`technique is still in its experimental stage with an error of
`1.25 mm in the final output that, obviously, needs to be
`reduced.
`
`OrthoCAD™ Technology
`
`OrthoCAD™ software has been developed by CADENT,
`Inc. (Computer Aided DENTistry, Fairview, NJ, USA)
`to enable the orthodontist to view, manipulate, measure
`and analyze 3D digital study models easily and quickly
`(Figures 3–5). Alginate impressions of the maxillary
`and mandibular dentitions, together with a bite registra-
`tion are required for the construction of 3D digital study
`models, which are then downloaded manually or auto-
`matically from the worldwide website using a utility
`called OrthoCAD Downloader. The average file size for
`each 3D model is 3 Mb.
`The operator can browse and view the models sepa-
`rately and together from any direction and in any desired
`magnification on screen (Figure 3). The software comes
`with several diagnostic tools such as: measurement analy-
`ses (e.g. Bolton analysis, arch width and length analyses);
`midline analysis (the ability to split the model sagittally
`or transversely for better comparisons); and overbite and
`overjet analyses (Figure 4). Any slight inaccuracies in bite
`registration can be compensated for by a function in the
`software, which enables anteroposterior or transverse
`shifting of one jaw. One of the interesting features of the
`program is the ‘Occlusogram’ (Figure 4). It includes
`color-coded occlusal views of the upper and lower dental
`arches, which allow the orthodontist to visually assess the
`inter-occlusal contacts. In addition, the operator has the
`ability to save, print or send any view on the screen to a
`colleague (or even to the patient) as an email attachment.
`
`

`

`JO June 2004
`
`Features Section
`
`3D imaging in orthodontics
`
`157
`
`Figure 2 The future of orthognathic surgical planning according to Khambay et al.16 is dependent on the use of CT-based maps in conjunction
`with stereophotogrammetry.
`
`Recently, a utility has been added to the software,
`‘OrthoCAD Virtual Set-up’, which is based on the
`straight wire philosophy (Figure 5). The assumption is
`that all the teeth are connected to the archwires and any
`manipulation of tooth position is done under the wire/
`appliance constraints. The orthodontist needs to go
`through 7 steps to reach the final plan (virtual treatment;
`see Figure 5). OrthoCAD™ Bracket Placement System
`is another addition to the system, which enables the
`orthodontist to position brackets according to their
`
`planned positions in the virtual treatment. More infor-
`mation can be obtained from OrthoCAD’s website.17
`
`Align® Technology
`
`Align® Technology, Inc. developed the Invisalign
`appliance for orthodontic tooth movement in the USA
`in 1998. It is an ‘invisible’ way to straighten teeth into a
`perfect occlusion using thin, clear, overlay sequential
`appliances.
`
`

`

`158 M. Y. Hajeer et al.
`
`Features Section
`
`JO June 2004
`
`Figure 3 (a) The penta-view of OrthoCAD. The operator can browse and view the models separately and together from any direction
`and in any desired magnification on screen. (b) Lingual aspects of the upper and lower teeth can be clearly seen and assessed using
`OrthoCAD™ manipulation tools.
`
`

`

`JO June 2004
`
`Features Section
`
`3D imaging in orthodontics
`
`159
`
`Figure 4 (a) The Occlusogram illustrates the amount of inter-occlusal contacts using color-coded scales. (b) Overbite and overjet can be assessed
`accurately by splitting the model along the mid-sagittal plane. (c) In addition to midline analysis, splitting can be performed at any point and in any
`angle. (d) Measuring mesio-distal widths of teeth. (e) Space analysis. (f) Three measurements of arch widths in the lower dentition. These are just a
`few of the available diagnostic tools with OrthoCAD™ software.
`
`

`

`160 M. Y. Hajeer et al.
`
`Features Section
`
`JO June 2004
`
`Figure 5 The first step in OrthoCAD™ virtual set-up is to choose your preferred brackets, bands and wires from the available straight wire
`systems listed. Secondly, incisors need to be positioned, as well as the molars (if required). In the third and fourth steps, the orthodontist should
`slide maxillary and mandibular teeth into their proper positions or correct the positions of the brackets themselves to achieve better inter- and
`intra-arch relationships. Extractions can be simulated at this stage and the resultant space can be manipulated manually or automatically. In the
`fifth step, the sagittal inter-arch relationships should be double-checked, followed by evaluating the transverse relationships in step 6. Finally,
`molar position and jaw closure are adjusted to make sure that the correct form of treatment is chosen.
`
`

`

`JO June 2004
`
`Features Section
`
`3D imaging in orthodontics
`
`161
`
`Figure 6 (a) Initial treatment planning with patients’ photographs and radiographs are sent to Invisalign®
`laboratories. (b) Impressions are converted into positives (plaster models) and checked for quality. (c) In
`the laboratory, models are first coated with protective shells, and encased in a mixture of resin and a
`hardener. (d) After chemical setting, they become blocks of hardened resin with many plaster models
`inside. Each tray is placed in a destructive scanning machine. (e) Each 3D model is constructed from about
`300 2D scans. Graphic designers cut out each tooth and save it as a separate geometric unit. (f) Once the
`teeth are separated and re-assembled back into the arches, the designers create a final set-up of what the
`patient’s teeth will look like when the treatment is completed. (g) For each stereolithographic constructed
`model (which represents a treatment stage), a clear Invisalign® aligner is created by heat. (h) These aligners
`are trimmed, polished, cleaned and finally sent to the prescribing orthodontist.
`
`

`

`162 M. Y. Hajeer et al.
`
`Features Section
`
`JO June 2004
`
`The Invisalign process begins with the orthodontist
`making an initial diagnosis and mapping out a course
`of treatment. Then these are sent to Align® Technology,
`together with the patient’s radiographs, impressions of
`the patient’s teeth and an occlusal registration. In the
`data acquisition laboratory, models are converted into
`3D data through ‘destructive scanning’ machines. The
`destructive scanning machine removes paper-thin slices
`of about 0.003 inch with a digital camera taking a 2D
`scan after each slice. A computer stacks together around
`300 of these digital images to create a 3D model. These
`data are then sent to graphic designers who cut out each
`tooth and save it as a separate geometric unit. Once the
`teeth are separated and reassembled back into the arches,
`the designers create a final set-up of what the patient’s
`teeth will look like when the treatment is completed
`(Figure 6). The treatment is divided into a series of stages
`that go from the current condition to the desired end
`result. This simulation is then electronically delivered to
`the orthodontist for final quality approval, following
`which a series of dental models are constructed from
`photosensitive thermoplastic. These are used to fabricate
`the finished product: a series of clear Invisalign aligners
`(Figure 6).
`The patient is instructed to wear each aligner for
`approximately 1–2 weeks, and then to move forward to
`the next stage. The first university-based clinical study
`reported successful clinical results of subjects with vary-
`ing degrees of mild to moderate malocclusion treated
`by this means.18 Although, the manufacturing company
`claims that the appliance can be used to treat Class II and
`III sagittal discrepancies, as well as vertical and trans-
`verse discrepancies, more clinical studies need to be con-
`ducted to prove or disapprove such claims. It is obvious
`that the treatment procedures do not allow for continued
`eruption of teeth or significant arch changes during
`growth.18 Dental movements can be achieved with this
`system, but not basal orthopedic changes. Any major
`change of tooth morphology during the treatment phase
`(e.g. restorations or composite build-ups) can destroy the
`use of subsequent aligners. The technique may not fully
`take into account optimum root positions at the end of
`the treatment, thereby ignoring one of the key factors
`in achieving prolonged stability and function. The
`Invisalign® philosophy treats only the teeth (more specifi-
`cally only the crowns) without building into the virtual
`treatment equation the relationship between the teeth
`and the cranial base, the lips and other oro-facial soft
`tissues.
`
`Acknowledgements
`
`We would like to thank OrthoCAD™ and Invisalign® for
`their kind help with some of the illustrative materials in
`this paper.
`
`References
`
`1. Mah J, Bumann A. Technology to create the three-
`dimensional patient record. Semin Orthod 2001; 7: 251–7.
`2. Hajeer MY, Ayoub AF, Millett DT, Bock M, Siebert JP.
`Three-dimensional imaging in orthognathic surgery — the
`clinical application of a new method. Int J Adult Orthod
`Orthog Surg 2002; 17: 318–30.
`3. Ayoub AF, Wray D, Moos KF, et al. A three-dimensional
`imaging system for archiving dental study casts: a prelimi-
`nary report. Int J Adult Orthod Orthog Surg 1997; 12: 79–84.
`4. McGuinness NJ, Stephens CD. Holograms and study
`models assessed by the PAR (Peer Assessment Rating) Index
`of malocclusion — a pilot study. Br J Orthod 1993; 20:
`123–9.
`5. Available at: www.orametrix.com (accessed August 2003).
`6. Case C. A Practical Treatise on the Technics and Principles of
`Dental Orthopedia. Chicago: Case Company, 1908.
`7. Van Loon J. A new method in dento-facial orthopedia,
`Part 1. Dental Cosmos 1915; 57: 1093–101.
`8. Simon P. Fundamental principles of a systematic diagnosis
`of dental anomalies (transl. B. F. Lischer). Boston: Stratford
`Company, 1926.
`9. Baumrind S. Integrated three-dimensional craniofacial
`mapping: background, principles, and perspectives. Semin
`Orthod 2001; 7: 223–32.
`10. Xia J, Wang D, Samman N, Yeung RW, Tideman H.
`Computer-assisted
`three-dimensional surgical planning
`and simulation: 3D color facial model generation. Int J Oral
`Maxillofac Surg 2000; 29: 2–10.
`11. Nishii Y, Nojima K, Takane Y, et al. Integration of the
`maxillofacial three-dimensional CT image and the three-
`dimensional dental surface image. J Japan Orthod Soc 1998;
`57: 189–94.
`12. Terai H, Shimahara M, Sakinaka Y, Tajima S. Accuracy of
`integration of dental casts in three-dimensional models.
`J Oral Maxillofac Surg 1999; 57: 662–5.
`13. Okumura H, Chen LH, Tsutsumi S, Oka M. Three-
`dimensional virtual imaging of facial skeleton and dental
`morphologic
`condition
`for
`treatment planning
`in
`orthognathic surgery. Am J Orthod Dentofac Orthop 1999;
`116: 126–31.
`14. Chen LH, Chen WH. Three-dimensional computer-assisted
`simulation combining facial skeleton with facial morphol-
`ogy for orthognathic surgery. Int J Adult Orthod Orthog
`Surg 1999; 14: 140–5.
`15. Curry S, Baumrind S, Carlson S, Beers A, Boyd R.
`Integrated three-dimensional craniofacial mapping at the
`Craniofacial Research
`Instrumentation Laboratory/
`University of the Pacific. Semin Orthod 2001; 7: 258–65.
`16. Khambay BS, Nebel JC, Bowman J, Ayoub AF, Walker F,
`Hadley D. A pilot study: 3D stereo photogrammetric image
`superimposition on to 3D CT scan images — the future of
`orthognathic surgery. Int J Adult Orthod Orthog Surg 2002;
`17: 244–52.
`17. Available at: www.orthocad.com (accessed Aug 2003).
`18. Boyd RL, Miller R, Vlaskalic V. The Invisalign system in
`adult orthodontics: mild crowding and space closure. J Clin
`Orthod 2000; 34: 203–13.
`
`