Three-dimensional assessment of facial asymmetry: A systematic review.

For patients with facial asymmetry, complete and precise diagnosis, and surgical treatments to correct the underlying cause of the asymmetry are significant. Conventional diagnostic radiographs (submento-vertex projections, posteroanterior radiography) have limitations in asymmetry diagnosis due to two-dimensional assessments of three-dimensional (3D) images. The advent of 3D images has greatly reduced the magnification and projection errors that are common in conventional radiographs making it as a precise diagnostic aid for assessment of facial asymmetry. Thus, this article attempts to review the newly introduced 3D tools in the diagnosis of more complex facial asymmetries.

Facial symmetry is a state of balance in which both right and left halves of the faces are perfectly related and hence homologous in size, shape, and position. On the contrary, absence or lack of balance between the right and left hemiface affects the facial proportions and results in facial asymmetry.[1]

A mild degree of asymmetry is common in the human face and often does not require any treatment.[2] <3% of skeletal asymmetries are of clinically insignificant.[3] This mild asymmetry is partly masked by the soft-tissue envelope, dental compensation, and change in head posture.[45]

Haraguchi et al. reported that in minor facial asymmetry, the right hemiface is wider than the left hemiface with a chin deviation to the left side.[6] Severt and Proffit et al. reported 5%, 36%, and 74% frequencies of facial laterality in the upper, middle, and lower thirds of the face, respectively. The Lower third of the face is more prone to facial asymmetry may be because of the longer span of postnatal growth.

For patients with facial asymmetry, complete and precise diagnosis, and surgical treatments to correct the underlying cause of the asymmetry are significant. In day-to-day practice, orthodontists deal with so many morphological and functional facial asymmetries. Hence, an assessment of facial symmetry has become an essential part of their clinical examination.

Submento-vertex projections and frontal facial photos were used initially for measuring facial asymmetry. In 1930s, after the introduction of conventional posteroanterior (PA) cephalograms, it has been used extensively in orthodontic and orthognathic diagnosis and treatment planning of facial asymmetry. Because the PA cephalogram is a two-dimensional (2D) view of a three-dimensional (3D) object, may subject to distortion and projection error. This result in differences between the real and PA cephalometric measurements which may reduce the diagnostic quality have been well-documented in the literature.[7] Rotation of the head inside the cephalostat is another common error related to the capture of PA cephalometric films. Cook et al. (1980) found that 5° of head rotation within the cephalostat has resulted in an obvious reversal of the side in which asymmetry was already present. Although the disadvantages have been widely acknowledged, the PA cephalometry is still an important tool in the assessment of facial asymmetry and treatment planning.

Since the introduction of 3D imaging methods in orthodontics, errors caused by conventional methods such as distortion and magnification are greatly reduced and quantitative measurements of the anatomic structures of interest became possible.[8] Even though various methods to determine facial asymmetry on 3D data have been developed and verified, none of them have been accepted universally.

The development of computed tomography (CT), cone beam computed tomogram (CBCT), stereophotogrammetry, and laser surface scanning has opened a new era in the field of diagnosis of facial asymmetry.[9] In addition, newly developed 3D CT software enables 3D reconstruction and precise measurement of the craniofacial complex.[10] 3D images can provide accurate and more detailed information for the diagnosis and treatment planning of facial asymmetry by means of quantitative measurement and comparison between the right and left sides of the structures. These 3D images can easily rotate and viewed from any angle. This rotating function facilitates us to analyze specifically the asymmetrical facial structures and to visualize clearly structures that cannot be well-documented with PA cephalometry. The aim of this article was to present a systematic review of the literature on the 3D assessment of facial asymmetry, which needs more attention in the present day orthodontic practice.

From the Two to Three Dimensions

The shortcomings of conventional 2D methods limit the therapeutically options related to facial asymmetry leading to an erroneous treatment. Development of 3D imaging methods in orthodontics took away the difficulties of conventional techniques and opens a broad area in the field of diagnosis and treatment planning.[11] Grayson et al. proposed a method of combining the PA and lateral cephalometric radiographs for the creation of a 3D grid to evaluate patients with craniofacial abnormalities like hemifacial microsomia. This “3D cephalogram” was advantageous in that normative cephalometric values could be utilized, however, it was such a time-consuming procedure and did not incorporate the soft-tissue deficiencies. This method was also inefficient to correct the problems frequently associated with 2D PA and lateral cephalometric radiographs.

The advent of the digital image to the orthodontics has provoked new research fields focused mainly on the diagnostic ability of the radiography through image manipulation. However, many digital images endure limitations similar to those experienced from the conventional radiographic methods including magnification, superimposition, and distortion of anatomical structures.

The 3D imaging techniques included: Morphanalysis,[12] laser scanning,[13] 3D CT,[14] stereolithography,[15] 3D ultrasonography,[16] 3D facial morphometry,[17] digigraph imaging,[18] moire topography,[19] and contour photography.[20]

With the introduction of the CT, the totality of the craniofacial complex can be assessed with high accuracy. This technology has become one of the most extensive in the image diagnostic methods nowadays. It is a fact that the CT has become reputable concurrently with the improvement of the scanners used. In 1977, Herman and Liu were the authors who introduced 3D reconstructions from axial slides which enhance the accuracy and reliability of 3D CT images.

The CT permits the exact determination and visualization of both the skeletal structures and soft tissues paradigm in 3D without the superimposition of anatomical structures.[2122] Unfortunately, the effective radiation exposure is higher in medical CT scanner than the conventional X-ray source. In addition, the CT is comparatively more expensive, and the scanners are not easily accessible.

The current scenario in dentistry for 3D imaging has moved away from conventional CT scans toward CBCT. CBCT has gained widespread popularity over CT since its introduction to dentistry in 1998, because of its similar quality but significantly less radiation exposure and cost effectiveness. The complex nature of the craniofacial anomalies demands the use of CBCT for the precise diagnosis and treatment planning. The CBCT has a 100 times lower radiation dose than that given by a medical CT scanner and even low radiation dose than produced by a periapical complete series. Moreover, conventional images such as panoramic, lateral, and anterior/posterior radiographs can be directly obtained from the CBCT.

The replacement of the conventional X-ray machines by the CBCT for the precise assessment of facial asymmetry is a potential advancement in the field of orthodontic diagnosis.

The following 3D techniques are frequently used to assess facial asymmetry with high precision

Stereophotogrammetry[23]

Although skeletal asymmetry exists, it may be masked by the overlying soft tissues or a normal face may appear asymmetric only because of the asymmetry of soft-tissues. Hence, the assessment of soft-tissue morphology is an inevitable part of the assessment of facial asymmetry. The most common category of 3D surface imaging system is relied on the digital stereophotogrammetric method. These systems are skilled to reproduce exactly the surface morphology of the face, and map realistic color and surface data onto the geometric figure resulting in a lifelike rendering.

The distances and angles can be calculated within a 3D coordinate system in stereophotogrammetry. In this technique, two photographs were captured with the help of two semi-metric cameras form a stereo pair. The 3D image of the object can be reconstructed with the use this stereo pair and analytical plotter.

In this method, a reference plane and two points (bilateral) are defined in space in such a way that the reference plane should be between and perpendicular to the line connecting these bilateral landmarks. The points should be liable to move to create a symmetrical arrangement to the reference plane. The minimal amount of movement needed by the bilateral landmarks to achieve a symmetric arrangement represents the asymmetry of the original configuration Dtotal. Exocanthion, endocanthion, superalare, and cheilion were the four reference planes used in this technique. The reference plane with the lowest Dtotal would be the preferred one to assess facial asymmetry in 3D, which is the plane perpendicular and bisecting the line connecting the bilateral landmarks.

Dynamic Modeling

Three-dimensional dynamic models have developed by Langenbach and Hannam[24] from the University of Columbia, Britan which can be used to analyze the relationship between structure and function. The models can be used to predict muscular, occlusal, and articular biomechanical events which take place during simulated function and to examine any of their deviations in form and function.

A dynamic 3D model is entirely different from the 3D data obtained from other 3D methods because in this technique all the functional matrices can be manipulated and changed individually and creates a true interactive patient-specific simulation. Orthodontist's can use this dynamic, patient-specific 3D digital model to formulate rapid and accurate diagnoses of facial asymmetry.

Laser scanning system[2526]

The laser scanner is a soft-tissue scanner and is widely used for capturing the 3D facial soft-tissue morphology, which is clinically more reliable and reproducible method for assessing the facial asymmetry. It consisted of 2 Minolta Vivid VI900 high-resolution 3D cameras, which operate as a stereo pair with a reported manufacturing exactness of 0.1 mm. The cameras were placed at a distance of 1350 mm from the subjects. Multi-scan software controls the scanners, and the data coordinates were stored in a vivid file format. The information gathered was transferred to reverse modeling software for further analysis. The natural head position is suggested for scanning with a total scan time of approximately 7.5 s. Both right and left laser scans have been taken simultaneously. The left and right scans were aligned based on the areas of overlap of the faces and thus how the asymmetry could be determined.

Cephalometrics in three-dimensional

A combination of lateral and frontal cephalograms all together put under the title “3D cephalometrics” however, these approaches are not true 3D. For 3D assessment of the maxillofacial complex,[27] a combined use of frontal, lateral, and submento-vertex views were advocated by some clinicians. Hence, the assessment of facial asymmetry is quite quantitative in nature; a 3D analysis with 2D radiographs will not be validated properly.[28]

In contrary, Kusnoto et al. has advocated the use of 3D cephalometric, specific software, to improve the exactness of 3D measurements.[29] They argued that with 3D cephalometric, measurement errors were reduced to 1.5 mm for linear 3.5° for angular measurements, a customized computer program. However, a special facebow must be needed to diminish the head positioning errors, and radiopaque markers are necessary to minimize landmark identification errors when taking the radiographs. Since Kusnoto et al. used 20 markers in his study; this technique is not practical for everyday practice.

Computed tomography scans

Computed tomography scans are frequently used to obtain 3D information on craniofacial complexes[10] and thus leads to the proper diagnosis of facial asymmetry. Because of the high level of radiation and long procedure in a confined space, patients and clinicians are hesitant to use conventional CT.

The advent of spiral CT has resolved these concerns.[9] In spiral CT, the patient is translated concurrently through the continuous rotation of the source-detector assembly, hence raw projection data are acquired in a relatively short period of time. This eventually reduces the amount of radiation exposure.

Cone beam computed tomogram

Cone beam computed tomogram imaging has greatly replaced spiral CT in dentistry due to its considerable reduction in radiation, and its invention has made 3D imaging easily accessible to the orthodontist who enhances the quality and accuracy of diagnosis.[30] CBCT images produce very precise and accurate 3D images of the craniofacial region. CBCT develop a 1-to-1 image-to-reality ratio, which is essential for accurate detection of the underlying deformities.[31] Moreover, the advantages of CBCT imaging for the assessment of asymmetry are suggested in the literature.

Cone beam computed tomogram imaging offers more precise information regarding the features of mandibular asymmetry than conventional PA cephalograms.[32] Therefore, when a visible chin deviation is present, CBCT scan should be preferred.

The infraorbital and mental foramina, foramina transversarium of atlas, inferior hamulus, medial and lateral condyles of the mandible, dens axis, superior, and mid-clinoid processes are the most reliable and reproducible landmarks tested for use in CBCT.[33]

The Digigraph[18]

Digigraph is a synthesize of video imaging, computer technology, and 3D sonic digitizing. The superimposition of skeletal and facial images have made possible with digigraph system without the use of radiography. The cephalometric measurements produced by this method are comparable to that of cephalometric radiographs and it consisted of a head supporting device, a digitizing probe, video cameras to record the extra-oral and intra-oral photographs, four microphone arrays to trace the landmarks and a computer to process the information. This device uses sonic, digitizing electronics to record the cephalometric landmarks by lightly touching the sonic digitizing probe to the patient's skin.

These imaging systems have the advantage of providing reliable measurements without trailing the sight of facial balance and harmony. The orthodontist can analyze the facial asymmetry using this system without exposing to an ionizing radiation.

Two-dimensional versus Three-dimensional in Landmark Specification

The cephalometric norms and most of the anthropometric tables in existence today are in a 2D form. Fortunately, with advances in software engineering and computer technology, it became possible to perform these analyses in 3D. 2D, the cephalometric landmarks articulare (Ar) and gonion (Go) are on the lateral part of the face and are significantly away from the midsagittal plane, where menton (Me) is positioned. 3D, the gonial angle is formed by the intersection of two lines, one that connects the Ar to Go, and another from Go anteromedially to Me. In the sagittal plane the size of this 3D angle differs from its 2D projection.

The difference between the 2D and 3D measurements are fairly minimal and clinically insignificant if the angles are relied on the midsagittal plane like SNA. 3D SNB angle in a patient with a significant laterognathia differs slightly only from its 2D counterpart is a good example.

Another common dilemma in 3D cephalometry is of angles formed by planes. When facial asymmetry exists, the landmarks on both the right and left sides of the face are not symmetric. Hence, the two planes may diverge from each other in all three planes of space. In these circumstances, the measure of the angle connecting the planes is a compound of pitch, roll, and yaw and is altered from the usual pitch measurement of 2D cephalometry.

Reference Specification in Assessment of Asymmetric Patients

Comparisons with bilateral differences or indices,[34] as well as using mirror images,[35] have been documented recently to assess facial asymmetry. These methods are critically based on which reference planes are to be used in the assessment. In the era of the soft-tissue paradigm, the estimation of facial asymmetry with the midsagittal plane as the reference plane, which relies only on skeletal landmarks does not suit well for establishing treatment objectives.

Kook and Kim has proposed of setting a new method of a transverse reference plane in CBCT scans to evaluate facial asymmetry.[36] Although 3D imaging methods are accurate enough for craniofacial analysis, landmark identification is challenging because there are no accepted 3D explanation of the conventional 2D landmarks.[37] In order to overcome these difficulties, a new reference plane which is a tangent line in utmost contact with the bilateral orbital floors, where the contact points are termed orbitale in (3D Or-3D) or bilateral Z is discovered.[38] This decreases the time needed for landmark identification. However, it was not documented that bilateral Or-3Ds are symmetric in asymmetric patients; it can then be used as a horizontal reference plane in such patients.

Three-dimensional digital image diagnosis data undergo a process called reorientation, which involves the head position adjustment on software images. This reorientation procedure has a significant impact on the results attained using 3D digital imaging data, particularly in CBCT.

The midsagittal plane obtained from CBCT image reorientation is not the final facial midline always, and adjustments are often mandatory depending on the patient's habitual head posture and treatment requirements. These adjustments could be minimized if soft-tissue reference points are actively considered during the image reorientation stage. Henceforth inclusion of soft-tissue landmarks, particularly those around the eye in the assessment of facial asymmetry granted an effective means of instituting a more precise facial midline in 3D CBCT image reorientations.[39]

Conclusion

Both 2D and 3D images are useful for better understanding of asymmetrical structures. Although most patients with the asymmetric face are well-diagnosed by cephalometric radiographs, certain circumstances require 3D imaging analysis to capture more accurate information. By observing and accurately estimate the factors that contribute to chin deviation, 3D imaging analysis will figure out its cause more accurately and efficiently. While considering the cost effectiveness, a combination of PA cephalometry and a frontal view photograph could be an effective tool for determining the presence and the degree of facial asymmetry. The further application of 3D imaging methods is suggested when PA cephalometry does not offer the information necessary for a complete and precise diagnosis.