Studying Facial Morphologies in Different Populations
Three- dimensional (3D) imaging of the facial region and its supporting structures has become very popular over the last 20 years.1-3 Traditional two- dimensional (2D) modes for record-taking have been replaced by 3D entities and are being used to diagnose orthodontic malocclusions. Although frontal and lateral cephalometric radiographs, panoramic radiographs, and intra – and extra – oral photographs are still used, more emphasis has been placed on the 3D virtual outlook4,5 and soft tissue esthetics.6 In the past, study models have probably been the leading “3D” records routinely used by practicing orthodontists and have given clinicians the ability to examine malocclusions from many viewpoints. Current virtual technologies have certainly enhanced the digitization of 3D models and have added value to the clinician.7
The paradigm shift in treatment philosophies also means that many clinicians have started to plan from the external profile, placing importance on the soft tissues of the face as they largely determine the limitations of orthodontic treatment. From the perspectives of function, stability, and esthetics, the orthodontist must plan treatment within the patient’ s limits of soft tissue adaptation and soft tissue contours.8 With the developments in technology, many clinicians have shifted towards digital computer- based records for quicker results, easier organization, the ability to enlarge and enhance images, and ease of sharing this information with patients and colleagues.
3D images of the facial soft tissue can help provide the clinician with this same degree of information with a more accurate representation of facial morphologies9-11 and can be useful to better understand, compare,3,12 and predict treatment outcomes before and after orthodontic treatment. 13–15 In addition, some 3D soft tissue models have been used to estimate growth changes.16,17
Some of the applications of 3D imaging in orthodontics include pre – and post – orthodontic assessment of dentoskeletal relationships and facial esthetics, auditing orthodontic outcomes with regard to soft and hard tissues, 3D treatment planning, and 3D soft and hard tissue prediction. 3D – fabricated custom – made arch wires and archiving 3D facial, skeletal, and dental records for in-treatment planning, research, and other medicolegal purposes are some other benefits of using 3D models in orthodontics.18
This chapter reviews the use of 3D imaging in the study of the facial morphologies of subjects from different populations.
We have known for some time that faces are different, especially when ethnic populations are involved. This is best illustrated using surface shapes and volumes from different populations. In this chapter, five different populations are discussed. Subjects were recruited from five sites in five different countries (Hungary [Hun], United Kingdom (Wales) [Wel], United States of America (Houston, Texas) [Hou], Slovenia [Slo], and Egypt [Egy]). All subjects were invited to participate in the study if they met the following inclusion criteria:
- ethnic descent/native of their country or state;
- between the ages of 18 and 30 years old;
- normal Class I malocclusions with no adverse skeletal deviations;
- normal body mass index values;
- no gross craniofacial anomalies.
Imaging s ystems
Two main imaging systems, the Minolta VI – 900 (laser scanning) and 3dMDface (stereo – photogrammetric) systems (Konica Minolta, Tokyo, Japan), were used in this study.
The laser scanning system consisted of two high- resolution Minolta VIVID VI- 900 3D cameras, with a reported manufacturing accuracy of 0.1 mm, operating as a stereo- pair. Each camera emits an eye- safe Class I laser, 690 nm at 30 mW, with an object – to – scanner distance of 600–2500 mm (depending on the lens type) and a fast mode scan time of 0.3 seconds. The system uses a one – half – frame transfer charge – coupled device and can acquire 307,000 datapoints. The scanner’ s output data is 640 × 480 pixels for 3D and red, green, and blue color data. Data were recorded on a desktop workstation, and for surface capture a Minolta medium- range lens with a focal length of 14.5 mm was used. The cameras were placed 1350 mm from the subjects. The scanners were controlled with multiscan software (Cebas Computer, GmBH, Eppelheim, Germany), and data coordinates were saved in a software file format known as a VIVID file format (.vvd). The system has previously been validated.19,20
The 3dMDface system is a structured light system using a combination of stereo- photogrammetry (a technique used to acquire 3D objects from stereoscopic images) and the structured light technique, and was also a portable system. 13 This system uses a multicamera configuration, with three cameras (one color and two infrared) on each side that capture photorealistic- quality pictures. A random light pattern is projected onto a subject, and an image is captured with multiple synchronized digital cameras set at various angles in an optimum configuration. This system is able to capture full facial images from ear to ear and under the chin in 1.5 milliseconds at the highest resolution. The manufacturer’ s accuracy is < 0.5 mm, and the quoted clinical accuracy is 1.5% of the total observed variance.21 3D surface images captured by surface acquisition systems are highly repeatable, and 3D landmark data can be acquired with a high degree of precision.2,22
Image a cquisition
In order to produce an accurate facial structure, natural head posture was adopted for all subjects, as this has been proven to be clinically reproducible.23 The subjects sat on an adjustable chair and were asked to look into a mirror with a horizontal and vertical line marked on it. They were asked to level their eyes to the horizontal line and to adjust the midline of their faces to line up with the vertical line. Adjustments to seating heights were made to assist the subjects in achieving natural head posture. The subjects were asked to swallow hard and to keep their jaws in a relaxed position just before the images were taken. Each image acquisition took 1.5 milliseconds.
Processing of f acial s hells
Completed 3D facial images were then imported into a reverse modeling software program (Rapidform 2006, or RF6; INUS Technology, Inc, Seoul, South Korea) for analysis. Together, these functions allowed high – quality polygon meshes, accurate freeform nonuniform rational B – spline surfaces, and geometrically perfect solid models to be created. RF6 generated data as absolute mean shell deviations, standard deviations of errors during shell – to – shell overlaps, maximum and minimum range maps, histogram plots, and color maps.
The initial file formats imported into RF6 had a semi- rough image texture due to the irregularity of the surface contours and the way in which light was reflected off the surfaces of different objects. Further data- processing by a custom- built software subroutine to produce a workable image that preserved the shape, surface, and volume.24 The images were checked individually, and unwanted areas that could not be automatically removed were removed manually by dividing the unwanted areas from the main shell before proceeding to the next stage. Surface meshes with “defects” or “holes” were filled-in automatically by RF6. Finally, one composite whole face per individual subject was generated.
Average f ace c onstructions
Average faces were constructed in this study to represent the average, and their variations were based on the facial morphology of different study populations. The average facial constructs were achieved by using a previously validated software subroutine created from tools available within RF6. The steps required to produce an average face have been discussed previously25 in this book.
Parameters m easured
Ten average male and female facial shells from each population were created and established. Each gender-specific average face was superimposed onto a population- specific gender template, using a specialized superimposition technique26 to compare morphologic differences between each one. This was done until all possible combinations were achieved. The method of superimposition utilized a systematic process and involved manually aligning five points of the facial scans (2 points on the inner canthus of the eyes, 2 points on the outer commissure of the lips, and 1 point on the nasal tip) and subsequently fine registration in which the computer software RF6 determined the best-fit of the two scans.
The following areas were analyzed:
- linear measurements
- color histograms
- surface areas/shapes.
Linear measurements representing the mean differences between two surface shells were measured and recorded in millimeters. This va/>