The objective of this study was to compare the accuracy of 3-dimensional (3D) digital facial photographs taken by the Bellus3D Face Camera Pro (Bellus3D) (Bellus3D Inc, Los Gatos, Calif) and the 3dMDface system (3dMD) (3dMD Inc, Atlanta, Ga) with one another and with direct anthropometry (DA).
A mannequin head was selected as the research object. Twenty facial landmarks were labeled on the basis of the 8 interlandmark distances and 5 angles that were defined. A 150-mm digital Vernier caliper (Mitutoyo Inc, Tokyo, Japan) with an accuracy of 0.02 mm was applied to directly measure the interlandmark distances, and the angles were calculated according to the law of cosines. All the measurements were conducted 3 times by each operator under identical conditions. Then, each scanner was used to acquire 3D photographs 5 times, generating 10 digital stereophotographs. Linear distances and angles were measured on the 3D facial photographs reconstructed with open-source MeshLab software (ISTI [Italian National Research Council], Rome, Italy). Each linear distance and angle were measured 3 times by 1 operator, and 3 examiners conducted the measurements independently. To obtain the trueness, equivalence tests were applied to compare the measurements of the 2 scanners and DA. In addition, the average absolute deviations were calculated to directly compare the trueness of 3 methods (Bellus3D vs 3dMD vs DA). Finally, the intraclass correlation coefficient was used to assess the interobserver agreement and the precision of 2 scanners.
As for the trueness, 7 out of 8 of the linear distance measurements (N-Pn, Sn-Pog, ORE-IRE, OLE-ILE, RLC-RMC, LLC-LMC, and CR-CL) and 3 out of 5 of the angular measurements (MLA, NFA, and INI) obtained by 3dMD were equivalent to those obtained by DA. Five out of 8 measurements (N-Pn, Sn-Pog, RLC-RMC, LLC-LMC, and CR-CL) and 1 out of 5 of the angular measurements (MLA) obtained by Bellus3D were equivalent to the measurements obtained with DA. All but 3 of the measurements (ORE-IRE, NFA, and INI) obtained with Bellus3D were equivalent to 3dMD. The mean absolute difference between 3dMD and DA was 0.36 ± 0.20 mm and 0.45° ± 0.56°; the deviation between Bellus3D and DA was 0.61 ± 0.47 mm and 0.99° ± 0.61°; and the deviation between Bellus3D and 3dMD was 0.38 ± 0.37 mm and 0.62° ± 0.39°. Regarding the precision of the 2 scanners, the intraclass correlation coefficient value of 3dMD was approximately 1.00, and that of Bellus3D was 0.99. The interobserver agreement for each linear and angular measurement was 0.99.
The trueness of each scanner was clinically acceptable for diagnosis and treatment planning. The precision of 3D photographs obtained by 3dMD and Bellus3D showed good scanning repeatability. The interobserver agreement between the 3 operators was rated as excellent (0.99).
High-resolution 3-dimensional images can today be easily generated using facial scanners.
The accuracy of 2 facial scanners was compared with direct anthropometry.
Images generated by 2 facial scanners were equivalent (deviation values 1 mm or 1°).
In the field of orthodontics, esthetic analysis carries great significance. Orthodontists often employ photography to assist in their esthetic diagnosis and document their patients’ pretreatment and posttreatment esthetic conditions. Two-dimensional (2D) photography is widely used to evaluate patients’ facial esthetics. , Although 2D cameras are simple to operate, they have notable limitations. For instance, measurements based on traditional 2D photographs are not accurate enough to record parameters of a patient’s stereo facial surface. , Direct anthropometry (DA) is an affordable and reliable method for measuring 3-dimensional (3D) surfaces, and it is historically regarded as the gold standard for facial measurements. Anthropometric analyses have been used in various scenarios for scientific purposes to characterize the faces of infants with cleft palates, study human mandibles, and observe the growth of the head. However, this classical method is too time-consuming to keep pace with modern demands for efficiency.
Three-dimensional photography, which combines several images of an object taken from different perspectives into one 3D topography, has been used in dentistry since 1922. Back then, the technology had obvious shortcomings such as technical limitations, low resolution, and prohibitively high prices. With the development of technology, high-resolution 3D images can be easily generated using the latest facial scanners. These facial scanners use 2 steps to generate a 3D facial photograph. First, the geometry, or shape data, of the patient’s face is presented by converting 2D facial data into a series of x-, y-, and z-definitions. Next, color texture information is applied to the shape data. The technologies used for generating 3D images include lasers, structured light, and both passive and active stereo photogrammetry. ,
In 3D surface imaging systems, the 3dMDface system (3dMD) (3dMD Inc, Atlanta, Ga), which incorporates passive and active stereophotogrammetry, has been widely used clinically. Its 3D image contains a 180° face capture (ear-to-ear), and its geometric accuracy, as claimed by the manufacturer, is 0.2 mm root mean square or better, depending on its operation mode. , The 3dMD consists of 2 modules, each with 6 machine vision cameras, and the shooting process does not require any stitching of the images. , However, its shortcoming is that it is too space-consuming for a private dental practice yet not portable enough for home operation. In recent years, a variety of low-cost digital 3D scanners have flooded the market. Among these newly released devices, the handheld Bellus3D Face Camera Pro (Bellus3D) (Bellus3D Inc, Los Gatos, Calif) caught our attention because it allows orthodontists to take 3D facial photographs of patients with a smartphone or tablet computer. With a price tag of $499, it is significantly less expensive than the 3dMD, which costs $20,000. The Bellus3D has a voice-guided user interface that allows patients to operate the unit themselves. The selfies taken by patients can not only be stored in their digital album but also can be exported and sent to an orthodontist in object (OBJ) format. This allows orthodontists to monitor a patient’s facial profile progress without the patient coming to the office, which is a simple and reliable method that is especially useful during the coronavirus disease 2019 pandemic. In addition, recent studies have shown that both the scanning time and data processing time of the Bellus3D are faster than units such as the EinScan Pro 3D scanner (Shining 3D Technology, Hangzhou, China) and Planmeca ProMax 3D Mid (Planmeca USA, Inc, Hoffman Estates, Ill), which suggests it merits further study. It is also worth mentioning that the Bellus3D is compatible with a wide variety of digital design software such as Dolphin Imaging (Dolphin Imaging and Management Solutions, Chatsworth, Calif) and Exocad (Exocad GmbH Inc, Darmstadt, Germany) to aid in treatment planning.
In contrast to previous studies only focusing on the performance of the Bellus3D, our study compares the trueness and precision of the Bellus3D and 3dMD with one another and with DA. We believe that our study will add new ideas and knowledge to the field of stereophotogrammetry.
Material and methods
A mannequin head was selected as the research object because it can stay completely still, and its facial expression will not change. Twenty facial landmarks were labeled in permanent red ink, as shown in Figure 1 , A . Using these landmarks, 8 linear and 5 angular measurements were defined as shown in Table I .
|N-Pn||Nasion, pronasale||The vertical linear measurement from nasion to pronasale|
|Pn-Sn||Pronasale, subnasale||The sagittal linear measurement from pronasale to subnasale|
|Sn-Me||Subnasale, menton||The vertical linear measurement from subnasale to menton|
|ORE-IRE||The outer end of the right eyebrow, inner end of right eyebrow||Transverse linear measurement of right eyebrow width|
|OLE-ILE||The outer end of the left eyebrow, the inner end of the left eyebrow||Transverse linear measurement of left eyebrow width|
|RLC-RMC||Right lateral canthus, right medial canthus||Transverse linear measurement of the right eye from exocanthion right to enocanthion right|
|LLC-LMC||Left lateral canthus, left medial canthus||Transverse linear measurement of the left eye from exocanthion left to enocanthion left|
|CR-CL||Cheilion right, cheilion left||Transverse linear measurement of mouth width|
|NFA||Glabella, nasion, pronasale (°)||The angular measurement from glabella to nasion to pronasale (nasofrontal angle)|
|NLA||Columella, subnasale, labium superior (°)||The angular measurement from columella to subnasale to labium superior (nasolabial angle)|
|MLA||Labium inferior, supramental, pogonion (°)||The angular measurement from labium inferior to supramental to pogonion (mentolabial angle)|
|NPS||Nasion, pronasale, subnasale (°)||The angular measurement from nasion to pronasale to subnasale|
|INI||Inner end of right eyebrow, nasion, inner end of left eyebrow (°)||The angular measurement from the inner end of the right eyebrow to nasion to the inner end of the left eyebrow|
A 150-mm digital Vernier caliper (Mitutoyo Inc, Tokyo, Japan) with an accuracy of 0.02 mm was applied to directly measure the 8 interlandmark distances mentioned above on the mannequin head. For the angular measurements, the Vernier caliper was applied to measure the 3 sides of a target triangle. For instance, to calculate the value of the nasolabial angle, 3 sides, including distances from columella to subnasale, from columella to labium superior, and from subnasale to labium superior, should be measured. The angles were calculated according to the law of cosines, and the formula for angle measurement is as follows:
All the measurements were conducted 3 times by each operator under identical conditions, and 3 independent operators (J.L, C.Z., W.L.) who were blinded to each other’s measurements participated in the measurements. The values of the 8 linear distances and the 5 angles measured by DA were taken as the true values, to which the values measured from the 3D reconstructions obtained by the 2 scanners were compared.
Two scanners were used to acquire 3D photographs: the Bellus3D and the 3dMD. The former consists of a Samsung Galaxy Tab 3 (Samsung Electronics, Seoul, South Korea) running the Bellus3D Face Camera app and Bellus3D external hardware with 3 cameras and 2 sensors ( Table II ). The 3dMD consists of 2 modules with 6 machine vision cameras per module, an industrial-grade flash system, and a stand. This scanner system incorporates both active and passive structured lights in stereophotogrammetry. , The 3D facial photograph outputs of both systems are in the OBJ file format, which includes information regarding surface texture and color ( Table III ).
|Operating system||Android 4.2 Jelly Bean|
|RAM memory||1 gigabyte|
|Flash memory||16 gigabytes|
|Camera||3 megapixels rear-facing, 1.3 megapixels front-facing|
|Dimension||188 × 111.1 × 9.9 mm (7.40 × 4.37 × 0.39 in)|
|Weight||306 g (10.79 oz)|
|Components||3dMD Face System||Bellus3D Face Camera Pro|
|Hardware||Two modules with 6 machine vision cameras per module; industrial-grade flash system; stand||Tablet computer accessory with 3 cameras and 2 sensors; tablet computer|
|Imaging modality||Hybrid passive and active structured light||Active structured light (infrared)|
|Acquisition time||~1.5 ms||15-25 s|
|Background requirements||Dark||Natural light|
|Operator||Experienced||Experienced and inexperienced|
|Head moving||Not required||Required|
The mannequin head was scanned 5 times using each scanner (for a total of 10 times) according to the manufacturer’s protocols. The image acquisition process was as follows: (1) before each capture, calibration was carried out in accordance with the instructions. After calibration, the mannequin head, fixed on a tripod with acrylic resin, was positioned before the 3dMD static camera. The mannequin head’s Frankfort plane was oriented parallel to the ground ( Fig 2 ). After capturing, automatic 3D reconstructions were supplied and then exported as OBJ files; and (2) the handheld Bellus3D scanner was operated by an experienced operator in a room with no windows and 1000 lux, which was recorded by a digital illuminance meter (AS803 Smart Sensor, Shenzhen, China). The mannequin head was kept still, with its Frankfort plane parallel to the floor. When scanning with the Bellus3D, the voice command guided the operator to move in a circular arc path around the model. The initial scanning position was in front of the mannequin head, and then the scanner was moved to the right side of the model. Next, it was moved from the right side to the left side and finally back to the front ( Fig 3 ).