Digital three-dimensional image fusion processes for planning and evaluating orthodontics and orthognathic surgery. A systematic review

Abstract

The three important tissue groups in orthognathic surgery (facial soft tissues, facial skeleton and dentition) can be referred to as a triad. This triad plays a decisive role in planning orthognathic surgery. Technological developments have led to the development of different three-dimensional (3D) technologies such as multiplanar CT and MRI scanning, 3D photography modalities and surface scanning. An objective method to predict surgical and orthodontic outcome should be established based on the integration of structural (soft tissue envelope, facial skeleton and dentition) and photographic 3D images. None of the craniofacial imaging techniques can capture the complete triad with optimal quality. This can only be achieved by ‘image fusion’ of different imaging techniques to create a 3D virtual head that can display all triad elements. A systematic search of current literature on image fusion in the craniofacial area was performed. 15 articles were found describing 3D digital image fusion models of two or more different imaging techniques for orthodontics and orthognathic surgery. From these articles it is concluded, that image fusion and especially the 3D virtual head are accurate and realistic tools for documentation, analysis, treatment planning and long term follow up. This may provide an accurate and realistic prediction model.

Facial soft tissue (skin, connective tissues, fat and muscles), facial skeleton (bone and cartilage) and dentition are the three important tissue groups in orthodontics and orthognathic surgery, which can be referred to as a triad . Together with other structures such as the superficial musculoaponeurotic system, the skeleton and dentition support the facial soft tissue surfaces. The triad plays a decisive role in planning orthodontic therapy and orthognathic surgery. Patients with a dysgnathic deformity need careful assessment of the facial soft tissues surface, the underlying maxillofacial skeleton and the dento-alveolar position and their interdependency.

Imaging and fusion techniques to analyse the facial profile, the facial skeleton and dentition for planning orthodontic therapy and orthognathic surgery have been available for over a century and can be described as analogue and digital techniques and image fusion models.

Analogue techniques . 10 years after the first orthognathic surgery for a congenital deformity was carried out , B abcock (1897) introduced the use of plaster casts for model surgery. This method of preoperatively performing the planned osteotomy on a dental plaster cast is still known as the ‘gold standard’ for planning postoperative occlusion. One decade later orthodontists start using anthropometry, clinical photographs, dental and facial plaster casts and early fusion models (1915–1926) for treatment planning. The development of these early fusion models was almost entirely abandoned in 1931 when Broadbent claims that cephalograms are more accurate for treatment planning, because they display the dentition in relation to the facial skeleton . Cephalograms were soon accepted as ‘the gold standard’ for planning orthodontic treatment and orthognathic surgery. In this way, clinicians started to concentrate on two of the three structures of the triad (facial skeleton and dentition), despite the fact, that the overlying soft tissues define the facial outline .

The disproportional focus on facial skeleton and dentition is evident in the treatment outcome of the patient, as this approach sometimes results in a good functional but poor aesthetic result. The use of cephalograms shows clinicians that some profile-related problems cannot be solved by creating a perfect dental arch with normal occlusion and that sometimes surgical displacement of the mandible and/or maxilla is required .

In the 1970s and 1980s, there was growing awareness that the aesthetic outcome is of equal importance to the patient as the rearrangement of the occlusion. Methods of studying the facial profile or for planning surgical treatment, with for instance Obwegeser’s ‘Wunschprofiel’ and methods of analysing the facial soft tissue surfaces were (re)introduced, including facial plaster casts , anthropometry and analogue photography .

Digital techniques . Digital photography was introduced to evaluate facial harmony . It allows clinicians to establish a more proportional focus on all three structures of the triad, to assess the patient’s deformity . An accurate and objective assessment of a facial deformity or a preoperative prediction of the surgical outcome in two dimensions, especially regarding asymmetry, will always be deficient since it does not address the volumetric changes of all the facial portions that determine neuromuscular balance and facial harmony. As a consequence, with a computer graphic two-dimensional (2D) representation of facial appearance , it is not possible to achieve a realistic and acceptable result. From the 1980s, the shortcomings of these techniques induced an increase in the use of three-dimensional (3D) imaging techniques , such as facial surface laser scanning , 3D stereophotogrammetry (3D photography) and (3D) video-imaging to render the facial soft tissue surface. Reconstructions of digital imaging and communications in medicine (DICOM) files from multislice CT (MSCT), cone-beam CT (CBCT) imaging or MRI slices to display the skeletal structures and digital dental models to display the dentition were also investigated ( Table 1 ).

Table 1
Imaging techniques for the facial soft tissue surface, the facial skeleton and the dentition.
Technique/hardware Advantages Disadvantages 3D digital 3D virtual head
Facial soft tissue surface
2D photography Accurate
Easy
Low costs
Not 3D
Volumetric (CB)CT data necessary to match textured surface
No Yes
MRI See below See below Yes No
3D ultrasonography Low costs
No ionizing radiation
Time-consuming
No textured surface
Deformation of soft tissue due to contact between probe and skin
Yes No
3D laser surface scanning Accurate data Harmful to eyes
Long acquisition time
Multiple scanners necessary for textured surface high costs
Sensitive to light and metal objects
Yes No
3D photography/3D stereophotogrammetry No radiation
Accurate and metrically correct data
Short acquisition time (2 ms)
Textured surface soft tissue
Low costs
Poor accuracy eye lenses
Poor accuracy of subnasal and submental area
Yes Yes
Facial skeleton
MSCT data High quality images Horizontal scanning position
high dosis ionizing radiation
High amount of streak artifacts
high costs
Out of office imaging
Yes No
(CB)CT reconstruction Upright scanning position
Reduced ionizing radiation
In office scanning
Acquisition time of 40 s or less
Relatively more noise in data
No Hounsfield unit calibration
Yes Yes
MRI No ionizing radiation
Accurate information of different layers soft tissue
High costs
Horizontal scanning position
No textured surface data
Long acquisition time
Yes No
Dentition
Digitized plaster cast Reduction of streak artifacts Plaster casts mandatory Yes Yes
(CB)CT reconstruction See above See above Yes No
CT/laser scanned impression Plaster casts not needed
Correct occlusion with wax bite
Possibility to produce a plaster casts remains
Impression is mandatory Yes Yes
Digital impression by intraoral scanning device No impression needed
Acquisition time 30 s
Easy to use
Patient friendly
Spray on dentition Yes No

FSTS, facial soft tissue surface; FS, facial skeleton; DT, dentition; (CB)CT, cone-beam computer tomography; MSCT, multislice CT; MRI, magnetic resonance imaging; DICOM, digital imaging and communications in medicine.

With CT data, it becomes feasible to produce a life-sized 3D milled model , a two stage resin model or a stereolithographic model of a patient. Various methods have been developed to integrate plaster casts into such models . The facial skeleton models allow the surgeon to analyse the patient’s deformity and plan orthognathic surgery in three dimensions. In such a 3D (augmented) model, model surgery can be performed only once and the soft tissue changes cannot be simulated. So although the third dimension is introduced, one of the structures of the triad (the facial soft tissues surface) is underestimated.

3D virtual planning software programs with a virtual operating room (VOR) were introduced at the end of the 1980s . The IT revolution (2000s) has enabled significant improvements of these software modules. . The reconstruction of DICOM files in a VOR enables the clinician to document, analyse and plan orthognathic surgery on a facial skeleton model as often and in as many different ways as required . Programs to analyse the facial soft tissue surface and dental models were also introduced. For the first time, these programs gave the clinician a true insight into all three structures of the triad, albeit separately and routinely on a 2D computer screen.

Since most of the 3D imaging techniques only display one of the three structures with optimal quality , it is evident that these imaging techniques are more powerful when they are used together. This emphasises the importance of image fusion of 3D image modalities to document and analyse the triad of a patient’s face accurately . This has enhanced a search for an ‘all in one’ assessment of the face. Such an assessment should be performed using one holistic digital data set as the result of an image fusion process, including the facial soft tissue surface, the facial skeleton and dentition: the 3D virtual head. This results in a realistic and accurate 3D fusion model, with the true rational relationships between the facial soft tissue surface, the facial skeleton and the dentition.

Image fusion models . An image fusion model is defined as a composition of at least two different imaging techniques. The principle of image fusion is based on the creation of a single data set that contains all three structures of the triad. With segmentation by thresholding it is possible to reconstruct a volumetric facial skeleton with dentition and an untextured 3D facial soft tissue surface . For example, a reconstruction of a (CB)CT contains the facial soft tissue surface representing the soft tissue, the bone volume representing the facial skeleton and the dentition, but the (CB)CT skin is untextured and the dental structures may contain streak artifacts caused by (in)direct restorations and/or orthodontic fixed appliances. To improve the quality of the virtual face and dentition, it is necessary to superimpose a textured facial soft tissue surface (e.g. acquired with a stereophotogrammetrical camera setup) and to upgrade or replace the dental images (e.g. with digital dental casts) .

3D data can be fused using three different methods : point based matching with or without the use of a reference frame; surface based matching ; and voxel based matching . The matching process of the first method is based on corresponding points, while the other two use congruent surface points or voxels (volumetric picture elements) of a manually selected region. Based on the triad, four possible 3D fusion models can be distinguished: image fusion of the facial skeleton and the dentition; image fusion of the facial soft tissue surface and the facial skeleton; image fusion of the facial soft tissue surface and the dentition; and image fusion of the facial soft tissue surface, the facial skeleton and the dentition.

Three methods are used to display the facial skeleton and the dentition: the life-sized stereolithographic (STL) or milled models augmented with dental casts; digital dental casts integrated in cephalograms ; and a 3D reconstruction of the (CB)CT with integrated digital dental casts . The first two are outdated. The third method virtually displays the facial soft tissue surface and the facial skeleton in 3D. The integration of digital dental casts into the CBCT reconstruction establishes an augmentation with improved visualisation of the dentition.

Apart from matching conventional or digital photographs with a lateral cephalogram, which is purely a 2D technique, three methods are used to fuse the facial soft tissue surface and the facial skeleton: matching a 3D photograph with a lateral and anteroposterior cephalogram ; mapping 2D photographs onto CBCT or MSCT data ; and fusing a 3D photograph or a 3D surface laser scan with the reconstruction of MSCT or CBCT data .

Several methods were developed about 80 years ago to display both the facial soft tissue surface and the dentition , which have all been abandoned. Nowadays, it is possible to fuse 3D data of the facial soft tissue surface with a digital dental model .

The integral fusion model consists of a (CB)CT reconstructed bony volume, in which the dental structures are replaced by a digital dental model and the textured facial soft tissue surface is superimposed upon the untextured facial soft tissue surface of the (CB)CT. This model visualises the textured facial surface, as well as the 3D skeletal structures and the dentition without artifacts .

Despite progress in surgical outcome (functionality, aesthetics and stability), orthodontists and oral and maxillofacial surgeons have not been able to develop an objective method to evaluate the soft tissue changes caused by orthognathic surgery , nor to predict the surgical outcome. To date, advanced 3D imaging techniques are available that can display the individual structures of the triad quite accurately, but none of the available craniofacial imaging techniques can capture the complete triad at once with optimal quality . It is thought that this can only be established based on a 3D image fusion process.

The aim of this systematic review is to summarise the state-of-the-art of 3D imaging and 3D fusion models in orthodontics and orthognathic surgery.

Methods and materials

Search strategy

The PubMed databases (Medline database, free open access of PubMed central, out of range articles, articles marked as ‘epub ahead of print’ and free full text articles) (1950 to 11 June 2009), and the OVID databases (Embase, 1980 to June 2009 and the Cochrane databases (the Database of Abstracts of Reviews of Effects and the Central Register of Controlled Trials)) (4 June 2009) were searched. No language limit was applied.

Four subqueries were defined, categorising the fusion processes, the head, 3D and the medical field of interest. The subqueries were combined for the overall systematic search in PubMed and OVID, which resulted in the following query: (generic OR image OR fusion OR fusio* OR registration OR registrated OR dataset OR augmented OR model OR superimposition* OR ‘composite model’ OR simulation) AND (head OR skull OR face OR facial) OR maxillofacial OR craniofacial OR craniomaxillofacial OR orofacial OR dentofacial OR hard tissue OR ‘hard tissue’ OR bone OR bony OR soft tissue OR ‘soft tissue’ OR virtual head OR ‘virtual head’ AND (3D OR 3-D OR ‘three dimensional’ OR three-dimensional OR three-dimensional imaging) AND (orthodontic* OR oral surgical procedures OR orthognathic* OR dysgnathic*).

Inclusion and exclusion criteria

As the first step, articles concerning image fusion models of the head were included. The following exclusion criteria were applied: studies concerning implantology or head and neck oncology, to limit the review to the field of imaging in orthognathic surgery; and studies describing integration of dental models into a stereolithographical 3D model, even though the resulting real model can be manipulated, the digital data cannot be manipulated.

As the second step, all articles discussing a fusion of at least two different 3D imaging techniques were included. Exclusion criteria were: measurements made on linked anteroposterior and lateral cephalograms, often referred to as 3D cephalometry (this form of cephalometry is not performed on a 3D image so models using it to register a digital dental cast or a 3D photograph were excluded); articles concerning navigation; and studies on prediction/simulation models, since these focused on marker registration or simulation algorithms, respectively, and did not discuss the image fusion processes itself.

The reference lists of each selected publication were hand-searched to complete the search, resulting in two additional articles.

Results

15 articles met the inclusion criteria. The QUORUM diagram is shown in Fig. 1 . All 15 articles described a fusion model of at least two different 3D digital imaging techniques. An overview of the characteristics of the included studies is presented in Table 2 .

Fig. 1
QUORUM flow diagram. Flow chart of the article selection process.

Table 2
Overview of fifteen articles discussing a 3D fusion model, including a complete overview of the integral fusion model.
Author Facial soft tissue surface Facial skeleton Dentition Data set for reconstruction Registration Bite registration In vivo / in vitro Patients ( N )
Image fusion of the facial soft tissue surface and the facial skeleton
A youb 3D photography MSCT DICOM Surface based In vivo (dentofacial deformities) 6
G roeve 3D photography MSCT DICOM Surface based In vivo (facial asymmetry) 1
K hambay 3D photography MSCT DICOM Surface based In vivo 1
M aal 3D photography CBCT DICOM Surface based In vivo , dysgnathic patients 15
Image fusion of the facial skeleton and the dentition
G ateno CT Laser scanned impression Fudicial markers Triple tray In vitro dry skull 1
G ateno CT Laser/CT scanned dental cast DICOM Fudicial markers Bite jig with fiducial markers In vivo (craniofacial deformities) 5
N kenke MSCT MSCT Scanned dental cast DICOM Fudicial markers Acrylic wafer In vivo 1
S chutyser CT CT scanned dental cast DICOM Point based Splint with gutta percha markers In vitro (dry skulls) 10
S wennen CBCT CBCT of triple tray impression DICOM Voxel based Wax bite In vivo (dysgnathic) 10
S wennen MSCT MSCT scanned dental cast DICOM Point based Acrylic wafer In vitro (dry cadavers) 10
S wennnen MSCT MSCT scanned dental impressions DICOM Point based Wax bite In vivo 10
S wennen CBCT CBCT scanned alginate impressions DICOM Surface based Modified wax bite wafer In vivo 10
U echi MSCT Laser scanned dental cast DICOM Fudicial markers Splint with fiducial markers In vivo (dysgnathic patients 2
Image fusion of the facial soft tissue and the dentition
R angel 3D photography CT scanned impressions Surface based In vivo 1
Image fusion of the facial soft tissue surface, the facial skeleton and the dentition
O lszewski MRI MSCT Laser scanned dental cast DICOM Wax bite In vivo 1
3D FIRG model a 3D photography CBCT Alginot triple scan DICOM Voxel and surface based Alginot impression In vivo
CBCT, cone-beam CT; MSCT, multislice CT; MRI, magnetic resonance imaging; DICOM, digital imaging and communications in medicine.

a Integral fusion model preferred by the authors (unpublished).

Four articles discussed a fusion model concentrating on the facial soft tissue surface and the facial skeleton. Nine concentrated on fusion models between the facial skeleton and the dentition, of which three studies were performed in vitro on skulls. One study was identified on a fusion model between the facial soft tissue surface and the dentition. One discussed a fusion model concerning the facial soft tissue surface, hard tissue and the dentition.

Results

15 articles met the inclusion criteria. The QUORUM diagram is shown in Fig. 1 . All 15 articles described a fusion model of at least two different 3D digital imaging techniques. An overview of the characteristics of the included studies is presented in Table 2 .

Fig. 1
QUORUM flow diagram. Flow chart of the article selection process.

Table 2
Overview of fifteen articles discussing a 3D fusion model, including a complete overview of the integral fusion model.
Author Facial soft tissue surface Facial skeleton Dentition Data set for reconstruction Registration Bite registration In vivo / in vitro Patients ( N )
Image fusion of the facial soft tissue surface and the facial skeleton
A youb 3D photography MSCT DICOM Surface based In vivo (dentofacial deformities) 6
G roeve 3D photography MSCT DICOM Surface based In vivo (facial asymmetry) 1
K hambay 3D photography MSCT DICOM Surface based In vivo 1
M aal 3D photography CBCT DICOM Surface based In vivo , dysgnathic patients 15
Image fusion of the facial skeleton and the dentition
G ateno CT Laser scanned impression Fudicial markers Triple tray In vitro dry skull 1
G ateno CT Laser/CT scanned dental cast DICOM Fudicial markers Bite jig with fiducial markers In vivo (craniofacial deformities) 5
N kenke MSCT MSCT Scanned dental cast DICOM Fudicial markers Acrylic wafer In vivo 1
S chutyser CT CT scanned dental cast DICOM Point based Splint with gutta percha markers In vitro (dry skulls) 10
S wennen CBCT CBCT of triple tray impression DICOM Voxel based Wax bite In vivo (dysgnathic) 10
S wennen MSCT MSCT scanned dental cast DICOM Point based Acrylic wafer In vitro (dry cadavers) 10
S wennnen MSCT MSCT scanned dental impressions DICOM Point based Wax bite In vivo 10
S wennen CBCT CBCT scanned alginate impressions DICOM Surface based Modified wax bite wafer In vivo 10
U echi MSCT Laser scanned dental cast DICOM Fudicial markers Splint with fiducial markers In vivo (dysgnathic patients 2
Image fusion of the facial soft tissue and the dentition
R angel 3D photography CT scanned impressions Surface based In vivo 1
Image fusion of the facial soft tissue surface, the facial skeleton and the dentition
O lszewski MRI MSCT Laser scanned dental cast DICOM Wax bite In vivo 1
3D FIRG model a 3D photography CBCT Alginot triple scan DICOM Voxel and surface based Alginot impression In vivo
Only gold members can continue reading. Log In or Register to continue

Feb 7, 2018 | Posted by in Oral and Maxillofacial Surgery | Comments Off on Digital three-dimensional image fusion processes for planning and evaluating orthodontics and orthognathic surgery. A systematic review
Premium Wordpress Themes by UFO Themes