Validation of In Vivo Assessment of Facial Soft-Tissue Volume Changes and Clinical Application in Midfacial Distraction: A Technical Report Emeka Nkenke, M.D., D.M.D., Astrid Langer, D.D.S., Xavier Laboureux, Dr. rer. nat., Michaela Benz, LAss., Tobias Maier, Dipl.-Min., Manuel Kramer, Gerd Häusler, Dr.-Ing., Ph.D., Peter Kessler, M.D., D.M.D., Ph.D., Jörg Wiltfang, M.D., D.M.D., Ph.D., and Friedrich Wilhelm Neukam, M.D., D.M.D., Ph.D. Erlangen, Germany
of visible volume changes by optical three-dimensional images can be carried out with considerable accuracy. The determination of volume changes and accompanying accommodation vectors completes the cephalometric analysis during the follow-up of patients undergoing midfacial distraction. The new parameters will help to assess normative soft-tissue data on the basis of three-dimensional imaging with a view to an improved three-dimensional prediction of the operative outcome of orthognathic surgery. (Plast. Reconstr. Surg. 112: 367, 2003.)
The purpose of this study was to validate the assessment of visible volume changes of the facial soft tissue with an optical three-dimensional sensor and to introduce new parameters for the evaluation of the soft-tissue shape achieved from three-dimensional data of selected cases of midfacial distraction. Images of a truncated cone of known volume were assessed repeatedly with an optical three-dimensional sensor based on phase-measuring triangulation to calculate the volume. Two cubic centimeters of anesthetic solution was injected into the right malar region of 10 volunteers who gave their informed consent. Three-dimensional images were assessed before and immediately after the injections for the assessment of the visible volume change. In five patients who underwent midfacial distraction after a high quadrangular Le Fort I osteotomy, three-dimensional scans were acquired before and 6 and 24 months after the operation. The visible soft-tissue volume change in the malar-midfacial area and the mean distance of the accommodation vector that transformed the preoperative into the postoperative surface were calculated. The volume of the truncated cone was 235.26 ⫾ 1.01 cc, revealing a measurement uncertainty of 0.4 percent. The injections of anesthetic solution into the malar area resulted in an average visible volume change of 2.06 ⫾ 0.06 cc. The measurement uncertainty was 3 percent. In the five patients, the average distance of maxillary advancement was 6.7 ⫾ 2.3 mm after 6 months and 5.4 ⫾ 3.0 mm after 2 years. It was accompanied by a mean visible volume increase of 8.92 ⫾ 5.95 cc on the right side and 9.54 ⫾ 4.39 cc on the left side after 6 months and 3.54 ⫾ 3.70 cc and 4.80 ⫾ 3.47 cc, respectively, after 2 years. The mean distance of the accommodation vector was 4.41 ⫾ 1.94 mm on the right side and 4.74 ⫾ 1.32 mm on the left side after 6 months and 1.62 ⫾ 1.96 mm and 2.16 ⫾ 1.52 mm, respectively, after 2 years. The assessment
The analysis of soft tissues of the face is an important part of planning orthodontic treatment and orthognathic surgery. The aim is to maximize function and aesthetics.1 The behavior of soft-tissue facial structures after distraction osteogenesis is especially of great interest. One of the prime aims of every orthognathic treatment is to attain a stable result. Therefore, in distraction osteogenesis, many reports have been dedicated to quantifying the postsurgical bony changes over the course of time. However, soft-tissue stability has not been studied as extensively.2 In this context, it is of special interest whether distraction histiogenesis can be found besides osteogenesis.3 For cephalometric analysis of soft and hard tissues preoperatively and postoperatively, lateral cephalograms and photographs are most
From the Departments of Oral and Maxillofacial Surgery, Orthodontics, and Optics, University of Erlangen-Nuremberg. Received for publication May 20, 2002; revised October 11, 2002. DOI: 10.1097/01.PRS.0000070720.66260.AE
367
368
PLASTIC AND RECONSTRUCTIVE SURGERY,
frequently used. However, quantitative assessment of the soft-tissue changes is confined to the midsagittal plane with both techniques.4 Therefore, the use of different techniques of three-dimensional imaging has been advocated to analyze soft-tissue changes away from the midline. Computed tomography, moire´ topography, stereophotogrammetry, and laser scanning techniques have been applied (Table I).4 –10 However, these techniques can lead to significant exposure to radiation, may harm retinal function, or require a time-consuming procedure of image acquisition.5 To overcome these shortcomings, phasemeasuring triangulation was used in the present study. Until now, optical threedimensional sensors based on this method have preferably been used for industrial measurement purposes. Now, the technique has been adapted to clinical applications. The aim of this study was to validate the assessment of visible volume changes with an optical three-dimensional sensor in vivo and to introduce new parameters for the evaluation of the facial soft-tissue shape based on threedimensional data. Moreover, the visible facial soft-tissue changes in five selected cases of midfacial distraction were followed for 2 years to implement this noninvasive, noncontact method of visualization of the facial surface into orthognathic surgery and to compare it to conventional cephalometric analysis.
tion method. A sequence of phase-shifted fringe patterns of structured light is projected onto the region of interest. The data are recorded from different directions by two charge-coupled device cameras. Subsequently, the images of the phase-shifted patterns are evaluated by means of a four-shift algorithm to receive the three-dimensional shape of the object’s surface. The three-dimensional sensor takes advantage of an astigmatic optical device for the projection of precise sinusoidal intensity-coded fringe patterns instead of a frequently used Ronchi-grating fringe projection. Because of low-pass filtering of the chargecoupled device camera and the defocusing of the imaging (illumination or observation), a sinusoidal pattern is interpolated much better than a Ronchi pattern.12 The switching between the phase-shifted fringe patterns is achieved by a light modulator (ferroelectric liquid crystal) during 16 camera video cycles. Because the astigmatic illumination optically performs an integration of the originally black-coded or white-coded pixels in one azimuth direction, the gray coding of the ferroelectric liquid crystal pixels is no longer needed to obtain a sinusoidal intensity-coded pattern. The sensor was calibrated for a measurement volume of 300 ⫻ 300 ⫻ 300 mm3, adapted to the average dimensions of the human head. The required measurement time for data acquisition is 640 msec. To allow the comparison of images of the same subject acquired at different times, they were superimposed and registered by the use of computer software (SLIM3D; 3D-Shape GmbH, http://www. 3D-shape.com/slim/slim_e.html). The three-dimensional views to be compared were registered by matching the regions of the facial surface that had not been altered by surgery. In the cases of midfacial distraction,
PATIENTS
AND
METHODS
Data Acquisition and Determination of Volume
An optical three-dimensional sensor (CAM3D; 3D-Shape GmbH, Erlangen, Germany, http://www.3D-shape.com/) was used for data acquisition. It was developed at the Department of Optics, University of ErlangenNuremberg (Fig. 1).11 The sensor is based on a modification of the phase-measuring triangula-
August 2003
TABLE I Previous Studies on Noncontact Soft-Tissue Analysis
Author
Method
Soft-Tissue Volume Determination
Validation of the Assessed Data
Aung et al., 19955 Bhatia et al., 19946 Bush and Antonyshyn, 19967 Ferrario et al., 19968 Kawano, et al., 19879 McCance et al., 19924 Motegi et al., 199910
Laser scanner Structured light scanner Laser scanner Infrared photogrammetry Moire´ stripes Computed tomography Laser scanner
No Yes No No No No Yes
Yes Yes Yes Yes No No No
Vol. 112, No. 2 /
369
MIDFACIAL DISTRACTION
pixel i is considered a small square with a surface ⌬x · ⌬y, whereby ⌬x and ⌬y are the sampling intervals in the x and y directions, respectively. The corresponding volume is given by the following formula:
⌬V i ⫽ ⌬x 䡠 ⌬y 䡠 z i,
(1)
whereby zi is the height at pixel i.The whole volume Vr of the considered region is computed as follows:
Vr ⫽
冘
i ⑀ region
⌬V i.
(2)
Finally, the difference in volume between the preoperative and the postoperative measurements is calculated as follows:
⌬V r ⫽ V r(postop) ⫺ Vr(preop).
FIG. 1. Optical three-dimensional sensor based on phasemeasuring triangulation (a, charge-coupled device camera; b, fringe pattern projector).
the supraorbital rim and forehead region were chosen for superimposition. The first step of the registration procedure is the manual removal of regions, which changed their shape postoperatively. Subsequently, the new postoperative data set is registered with the preoperative three-dimensional view. The two images are aligned interactively on the computer monitor by selecting three sets of corresponding points. Then, the remaining deviations are automatically minimized by an optimization algorithm. The relative shift and rotation parameters are calculated to fit preoperative and postoperative data sets in the same coordinate system. The matched preoperative and postoperative three-dimensional images are represented by color-coded visualization. Pixels of the corresponding image positioned in the foreground are visible. Parts corresponding to unchanged regions show a balanced mix between the two colors. To determine the registration error, the mean distance between preoperative and postoperative unchanged areas was assessed. To calculate the volume difference, the corresponding areas of the preoperative and postoperative images have to be isolated. Each
(3)
For the validation of the volume determination, optical measurements were performed on a reference object of known volume. A truncated cone was chosen, because its whole shape can be acquired with a single three-dimensional measurement without shadowing effects (Fig. 2). Three-dimensional data measurement of the truncated cone positioned on an even plane was performed. As a baseline measurement, an image of the even plane without the cone was assessed. The volume of the truncated cone was determined as the difference between the two data sets. The procedure was repeated 10 times. Ten volunteers received an injection (single-use syringe; B. Braun, Melsungen, Germany) of 2 cc of anesthetic solution (Ultracain D-S; Hoechst Marion Roussel, Bad Soden, Germany) in the left malar region. The design of
FIG. 2. Truncated cone for volume determination.
370
PLASTIC AND RECONSTRUCTIVE SURGERY,
the trial was chosen according to Bhatia et al. 6 It was approved by the Institutional Ethics Committee of the University of ErlangenNuremberg (approval no. 2221). At room temperature, the amount of the liquid was determined on microscales (CP323P; Satorius, Göttingen, Germany). Precisely 2.00 g of the solution was filled in the syringe corresponding with 2 cc. Before and immediately after the injections, three-dimensional images were assessed to calculate the visible volume increase. The measurement uncertainty of the volume determination of the truncated cone and the determination of the volume increase of the malar region after injection of anesthetic solution was expressed as the coefficient of variation [(SD/mean value) ⫻ 100; %].
termined clinically and cephalometrically on the basis of the severity of the midface retrusion and the anterior dental crossbite. All patients were observed weekly until the intended overbite was achieved. The device was kept for 2 months for rigid retention after activation was completed. Once the distraction device was removed, a Delaire mask was used for 2 additional months to prevent relapse. Lateral cephalograms and optical threedimensional images were obtained preoperatively (time 1), 6 months after the operation (time 2), and 2 years after the operation (time 3) (Orthophos CD; Siemens, Erlangen, Germany) (Fig. 3). Both were assessed with lips at rest. The position of the patient during the three-dimensional scanning procedure was adjusted reproducibly using a cephalometric head holder to prevent movement artifacts and soft-tissue distortion caused by altered inclination of the head. All radiographs were taken in the same cephalostat with the same distance between object and film. A correction was made to calibrate the lateral cephalograms, because they showed a magnification of 10 percent. The radiographs obtained at each interval were traced on acetate paper. Vertical and horizontal measurements were made of each tracing on an x-y coordinate system. The x coordinate was formed by a line rotated 7 degrees downward from the sella-nasion (SN) line. A perpendicular vertical reference line through the sella was drawn and used as the y coordinate. The x and y coordinates were drawn on the preoperative tracing of each patient. The subsequent tracings were superimposed on the original tracing and then measured on the x and y axes to maximize the consistency among measurements. Skeletal and dental landmarks were digitized to obtain the angular and linear measurements (Dentofacial Planner System; Gemetek Com, Erding, Germany).1 The following points were included: sella (S), nasion (N), anterior nasal spine (ANS), A point (A pt), prosthion (Pr), incision superius (IS), incision inferius (II), infradentale (In), B point (B pt), pogonion (Pg), gonion tangent point (Tgo), pterygomaxillare (Ptm), and a point on the line ANS-Ptm perpendicular to A (A'). The landmarks were used to determine the sella-nasion-subspinale (SNA) angle (average value, 82 degrees; lower values indicate a retroposition of the maxilla and higher val-
Patient Follow-Up
Five patients suffering from midfacial hypoplasia underwent maxillary distraction osteogenesis with advancement by use of submerged devices in August of 1999. Three female and two male patients underwent the surgical procedure. Their ages ranged from 9 to 19 years (Table II). The patients included two with unilateral cleft lip and palate, one with median cleft, one with ectodermal dysplasia with agenesis of teeth, and one with maxillary hypoplasia attributable to developmental reasons. All patients received preoperative and postsurgical orthodontics. A high quadrangular Le Fort I osteotomy in one or two pieces was performed. An intraoral distraction device (Zurich Pediatric Ramus Distractor; KLS Martin, Tuttlingen, Germany) was applied to the lateral aspect of the zygoma posterior to the vertical osteotomy and to the lateral maxilla anterior to the vertical osteotomy; 2-mm osteosynthesis screws were used for fixation.13 The activation rods perforated the vestibular mucosa of the maxilla. The latency period before maxillary distraction was 5 days. It was performed at a rate of 0.5 mm/day in two 0.25-mm increments. The duration of the activation period was deTABLE II Patient Data
Patient
1 2 3 4 5 Mean ⫾ SD
Date of Birth
Sex
Age at Operation (yr)
07-12-80 07-05-88 24-01-85 17-02-87 05-12-90
Female Male Female Male Female
19 11 14 12 9 13.0 ⫾ 3.8
August 2003
Vol. 112, No. 2 /
MIDFACIAL DISTRACTION
371
FIG. 3. Design of the patients’ trials. 3D, three-dimensional.
ues indicate an anteposition of the maxilla), the sella-nasion-supramentale (SNB) angle (average value, 80 degrees; lower values indicate a retroposition and higher values indicate an anteposition of the mandible), and the A pointnasion-B point (ANB) angle (average value for orthognathic subjects, 0 to 2 degrees; lower values indicate a retroposition of the maxilla or an anteposition of the mandible and higher values indicate a retroposition of the mandible or an anteposition of the maxilla), as well as the SN distance (length of the anterior skull base), the A'-Ptm distance (length of the maxilla), and the Tgo-Pg distance (length of the mandible). The optical three-dimensional images acquired before the operation (time 1), 6 months after the operation (time 2), and 2 years after the operation (time 3) were registered. The mean distances of the data points of the preoperative and the postoperative areas used for registration were assessed. Midsagittal sections, parasagittal sections through the middle of the globe, and horizontal sections through the middle of the roof of the nose were generated (Fig. 3). The midsagittal profiles were used for a cephalometric determination of soft-tissue landmarks. A line through orbitale (O) and porion (P) was used as the horizontal reference. The soft-tissue nasion (N'), the subnasale point (Sn), the superior labial sulcus (SLS), and the labrale superius (LS), labrale inferius (LI), sulcus inferior (SI), and soft-tissue pogonion (Pg') were assessed in the preoperative and postoperative data sets, distinguishing for sagittal and vertical dimension to determine the extent of change in pro-
file.1,14 The differences between the preoperative and postoperative landmarks were calculated. The soft-tissue–to– hard-tissue advancement ratio was determined for incision superius and labrale superius. Moreover, facial concavity (N'SnPg') was assessed. In the parasagittal and horizontal sections, the maximum advancement of the soft tissue was determined separately for both the left and the right sides (Fig. 4). The volume changes of the midfacial region were determined separately for both the left and the right sides. As borders for the volume determination, horizontal planes were placed through the labial commissure and 5 mm caudal to the most caudal aspect of the inferior eyelid. Vertical planes were positioned through the medial canthus and 5 mm preauricularly (Fig. 5, left). The mean distance of the accommodation vector that transfers the preoperative into the postoperative surface was determined for the right and left sides (Fig. 5, right). The soft-tissue–to– hard-tissue advancement ratio was assessed for incision superius and accommodation vector. At times 1, 2, and 3, the height and weight of the patients were documented. Occipitofrontal head circumferences were measured with a paper tape according to standard guidelines.15 For all cephalometric measurements, the data were acquired twice. The Dahlberg formula was used to determine the standard error for the variables in each data set.16 A standard error below 1.0 was considered to reveal sufficient accuracy of the measurements, keeping
372
PLASTIC AND RECONSTRUCTIVE SURGERY,
August 2003
FIG. 4. (Above, left) Registered facial surfaces (blue, time 1; green, time 2; red, time 3) with lines b-f indicating different planes of section. (Above, right, first row) Horizontal sections through the forehead corresponding to line b. (Below, left) Midsagittal sections corresponding to line c. (Below, center) Parasagittal sections right corresponding to line d. (Below, right) Parasagittal sections left corresponding to line e. (Above, right, second row) Horizontal section through the middle of the roof of the nose corresponding to line f.
in mind that there is a strong dependency of the standard error on the sample size. The reliability of the different measurements was checked by computing the variance of each pair of measurements for each patient. The variances were summed for the five patients
and the sum was divided by 5. Finally, the square root was taken to determine the mean intraindividual standard deviation of the measurements. All calculations were performed using SPSS for Windows version 10.0 (SPSS, Inc., Chicago, Ill.).
Vol. 112, No. 2 /
373
MIDFACIAL DISTRACTION
FIG. 5. (Left) Registered facial surfaces (region of interest marked by white rectangle; blue, time 1; green, time 2; red, time 3). (Right) Arrow indicates accommodation vector that transfers the preoperative surface into the postoperative surface.
RESULTS
The volume of the truncated cone was 235.26 ⫾ 1.01 cc, revealing a measurement uncertainty of the volume determination of 0.4 percent. After injection of 2 cc of saline solution in vivo, visible volume changes of 2.06 ⫾ 0.06 cc could be measured. The measurement uncertainty was 3 percent. The patients gained height during the follow-up (Table III). The body weight increased at time 3 (Table III). The head circumference changed minimally over the observation period (Table III). After the registration procedure of the three-dimensional images, the mean distance between forehead regions assessed at times 1 and 2 was 0.26 ⫾ 0.05 mm, whereas it increased to 0.38 ⫾ 0.08 mm on average after 24 months (Table IV). From the lateral skull radiographs, all intended hard-tissue landmarks could be deter-
mined. In the five patients, the length of the anterior skull base (SN) remained stable throughout the observation period (Table V). The average length of the maxilla (A'-Ptm) increased during the observation period (Table V). Preoperatively, SNA values below 80 degrees and negative ANB values revealed a retroposition of the maxilla in all patients (SNA, 75.46 ⫾ 3.52 degrees; ANB, ⫺4.28 ⫾ 2.79 degrees). At time 2, the SNA and ANB could be increased in all patients (SNA, 82.22 ⫾ 3.46 degrees; ANB, 2.46 ⫾ 1.06 degrees) (Table VI). At time 3, the average values showed a confined relapse (SNA, 80.42 ⫾ 3.11 degrees; ANB, 2.30 ⫾ 1.05 degrees) (Table VI). The incision superius was advanced 6.7 ⫾ 2.3 mm on average at time 2 and 5.4 ⫾ 3.0 mm at time 3, showing a slight tendency for relapse. There was no significant sign of relapse (Table
TABLE III General Growth Data
Head Circumference (cm)
Body Height (cm)
Body Weight (kg)
Patient
T1
T2
T3
T1
T2
T3
T1
T2
T3
1 2 3 4 5 Mean ⫾ SD
50.5 49.7 51.1 52.6 49.0 50.58 ⫾ 1.38
50.9 50.7 50.5 52.4 49.3 50.76 ⫾ 1.11
50.0 50.7 51.0 52.4 49.8 50.78 ⫾ 1.03
165 150 160 150 121 149.2 ⫾ 17.1
165 150 160 163 124 152.4 ⫾ 16.9
165 157 160 166 128 155.2 ⫾ 15.6
58 45 46 45 18 42.4 ⫾ 14.7
59 42 45 47 18 42.2 ⫾ 15.0
62 47 49 50 20 45.6 ⫾ 15.5
T1, time 1 (preoperative); T2, time 2 (6 months after operation); T3, time 3 (2 years after operation).
374
PLASTIC AND RECONSTRUCTIVE SURGERY,
TABLE IV Registration Error
Mean Registration Error (mm) Patient
1 2 3 4 5 Mean ⫾ SD
T2
T3
0.2 0.3 0.3 0.2 0.3 0.26 ⫾ 0.05
0.3 0.4 0.4 0.5 0.3 0.38 ⫾ 0.08
T2, time 2 (6 months after operation); T3, time 3 (2 years after operation).
VII). The overjet could be improved by 3.7 mm on average at time 2 (Table VII). At time 3, for the overjet in two patients, a relapse of 1.5 and 1 mm could be found (Table VII). From the three-dimensional facial surface images, all intended parameters could be assessed successfully. At time 2, a sagittal advancement of the labrale superius of 5.2 ⫾ 2.6 mm could be determined. At the end of the observation period, a slight relapse could be found (3.7 ⫾ 2.5 mm) (Table VIII). The facial convexity (N'SnPg') could be reduced from 184.0 ⫾ 10.1 degrees (time 1) to 165.2 ⫾ 2.5 degrees (time 2) (Table VIII). At time 3, the convexity remained stable (167.6 ⫾ 2.1 degrees) (Table VIII). Only one patient showed complete symmetry for the maximum visible soft-tissue advancement in a parasagittal and a horizontal plane at times 2 and 3. The maximal difference between the left and right sides was 5 mm at time 2 and 2 mm at time 3 (Table IX). The visible volume increase in the malarmidfacial region was larger on the left than on the right side at times 2 and 3 (righttime2, 8.92 ⫾ 5.95 cc; lefttime2, 9.54 ⫾ 4.39 cc; righttime3, 3.54 ⫾ 3.70 cc; lefttime3, 4.80 ⫾ 3.47 cc) (Table X). In one patient, a visible volume change in the malar-midfacial region of ⫺2.2 cc was determined after 2 years (Table X). The negative
August 2003
volume revealed that the soft-tissue volume had decreased below the preoperative value. At time 2, the mean distance of the accommodation vector showed only a minimal difference between right and left sides (rightvector, 4.41 ⫾ 1.94 mm; leftvector, 4.74 ⫾ 1.32 mm) (Table X). However, a pronounced reduction of the vector of more than 50 percent could be found at time 3 (rightvector, 1.62 ⫾ 1.96 mm; leftvector, 2.16 ⫾ 1.52 mm) (Table X). The ratio of advancement of the labrale superius to the incision superius was 77.4 ⫾ 11.0 percent at time 2 and 68.0 ⫾ 12.6 percent at time 3 (Table XI). The ratio of the accommodation vector to the incision superius at time 2 was 71.2 ⫾ 32.1 percent (right) and 79.2 ⫾ 35.8 percent (left) (Table XI). At time 3, the corresponding values were 47.0 ⫾ 80.5 percent (right) and 60.2 ⫾ 71.1 percent (left) (Table XI). For the different measurements, the determination of the standard error according to the Dahlberg formula revealed values below 1.0 for all variables, indicating sufficient accuracy of the measurements. However, it has to be kept in mind that the standard error shows a dependency on the sample size. The mean intraindividual standard deviations are given in Tables V through XI. DISCUSSION
The determination of three-dimensional changes after orthognathic surgery is often confined to two dimensions. Routinely, the soft-tissue analysis is based on lateral skull radiographs and standardized profile photographs of the patient. Quantitative evaluation of soft-tissue changes away from the midline is not performed.17 Several tools to evaluate the complete facial surface morphologically have been provided in recent years.8,18,19 Their use has been often limited to the study of small
TABLE V Growth of Anterior Skull Base and Maxilla
SN (mm) Patient
1 2 3 4 5 Mean ⫾ SD Mean intraindividual SD
A⬘-Ptm (mm)
T1
T2
T3
T1
T2
T3
58.2 69.6 61.6 68.5 58.6 63.30 ⫾ 5.42 0.11
56.6 67.1 61.9 67.5 57.1 62.04 ⫾ 5.23 0.14
58.9 70.3 62.5 70.1 58.0 63.96 ⫾ 5.94 0.13
43.6 44.8 43.3 41.5 37.2 42.08 ⫾ 2.97 0.27
43.8 45.1 42.9 46.7 37.4 43.18 ⫾ 3.53 0.10
43.8 47.0 49.8 48.0 35.4 44.80 ⫾ 5.69 0.11
T1, time 1 (postoperative); T2, time 2 (6 months after operation); T3, time 3 (2 years after operation).
T3
SNA, sella-nasion-subspinale (angle); SNB, sella-nasion-supramentale (angle); ANB, A point-nasion-B point (angle); T1, time 1 (postoperative); T2, time 2 (6 months after operation); T3, time 3 (2 years after operation).
2.3 4.3 2.1 2.0 1.6 2.46 ⫾ 1.06 0.04
T2 T1
–6.2 –2.8 –0.2 –7.1 –5.1 –4.28 ⫾ 2.79 0.04 81.1 75.3 89.0 81.1 79.5 81.20 ⫾ 4.96 0.20
T3 T2
80.1 72.1 85.9 80.3 82.4 80.16 ⫾ 5.07 0.21 81.2 73.8 85.8 82.2 80.7 80.74 ⫾ 4.36 0.11
T1 T3
83.3 76.0 83.2 78.7 80.9 80.42 ⫾ 3.11 0.14 83.2 76.4 85.4 82.2 83.9 82.22 ⫾ 3.46 0.10
T2 T1 Patient
75.0 70.9 80.8 75.1 75.5 75.46 ⫾ 3.52 0.16
ANB (degrees) SNB (degrees) SNA (degrees)
TABLE VI Angular Cephalometric Measurements
2.1 4.1 2.1 1.8 1.4 2.30 ⫾ 1.05 0.03
MIDFACIAL DISTRACTION
1 2 3 4 5 Mean ⫾ SD Mean intraindividual SD
Vol. 112, No. 2 /
375
groups of patients, because most of these techniques are substantially invasive or expensive.10 To overcome these shortcomings, the use of optoelectronic sensors seems to be reasonable. With short measurement times, they offer a noncontact, noninvasive technique without irradiation that permits analysis of facial changes away from the midline. The image is immediately available and can be viewed from varying directions using the appropriate shading. It is possible to store the image and retrieve it for superimposition. Although the equipment is specialized, it is of low cost compared with previously described devices, such as the laser scanner.20,21 A three-dimensional work station is not required. The expenses during clinical use are minimal. The aims of the present study were (1) to introduce a noncontact, noninvasive, optical, three-dimensional scanner based on phasemeasuring triangulation to oral and maxillofacial surgery, (2) to validate the measurements, and (3) to propose new parameters for the evaluation of the facial soft-tissue shape computed from three-dimensional data in cases of midfacial distraction. Only a few reports are available on the validation of volume assessments by optical sensors.6 They show errors of approximately 20 percent when in vivo measurements are taken to determine visible volume changes after the injection of 2 cc of anesthetic solution. With the same experimental design, the scanner used in the present study revealed measurements that produced an error of only 3 percent. Moreover, the measurements of the truncated cone showed the high potential of the method by a measurement error of only 0.4 percent. Currently, it seems that an optical method for volume determination with higher accuracy cannot be found in the literature. The follow-up interval was chosen according to the findings of previous reports. Several studies have suggested that the soft tissues attain definitive form after the sixth postoperative month.22,23 After 1 year, the results remain stable.2 Therefore, the follow-up examinations were performed 6 months and 2 years after operation. During the 2 years of follow-up, head circumference and length of the anterior skull base increased minimally or remained stable in the individual patients. Therefore, it seemed reasonable to use the forehead region for the registration of the three-dimensional images.
376
PLASTIC AND RECONSTRUCTIVE SURGERY,
August 2003
TABLE VII Hard-Tissue Advancement
Sagittal Advancement of Incision Superius
Overjet (mm) Patient
1 2 3 4 5 Mean ⫾ SD Mean intraindividual SD
T1
T2
T2
T2
T3
–9 –1 –4 –5 –7 –5.2 ⫾ 3.0 0.01
1 3.5 1 3 –1 1.5 ⫾ 1.8 0.01
1 2 1 2 –2 0.8 ⫾ 1.6 0.01
10 4.5 5 8 6 6.7 ⫾ 2.3 0.13
10 2 5 6 4 5.4 ⫾ 3.0 0.06
T1, time 1 (postoperative); T2, time 2 (6 months after operation); T3, time 3 (2 years after operation).
Registration errors were determined as mean distances between the forehead regions assessed at the different follow-up examinations. These distances are related to different aspects. Height and weight of the patients increased during the observation period. Although it has been found that facial growth in female subjects is largely completed by the age of 12 years, male subjects show growth until the age of 17.24 It has even been stated that adult enlargement continues in the craniofacial complex throughout life.25 Therefore, growth may have influenced mainly the registration error in the present study. However, for clinical applications, the submillimeter accuracy of the procedure seems adequate. Traditionally, attempts at comparing profiles with a view to quantifying changes resulting from surgical procedures have involved an operator identifying and locating landmarks.26 These landmarks are often sparse. Objections to their use have been raised on the grounds that the individual analysis is dependent on the profile orientation. The landmarks provide no information on the shape or the change of the shape in the segments joining them. Moreover, their determination relies on expert opinion to create homologous points.27 However, if landmarks are preferred, the three-dimensional images can be used for their determination. Distances and angles can be calculated in the three-dimensional data set. Errors related to soft-tissue distortion, which can occur during clinical measurement, can be avoided. In the present study, the standard landmarks for the profile analysis were determined from the three-dimensional data because they can be used to determine the midsagittal profile without uncertainties caused by summation effects or orientation failures during the assessment of lateral skull radiographs.
With the three-dimensional data, the individual user no longer has to rely on others when he or she cannot examine the patient personally to determine the landmarks. The data set virtually documents the state of the facial tissue at a well-defined point in time without loss of information. Although the patient is absent, the landmarks can be assessed on the virtual face according to the examiner’s interpretation of the definition of each single landmark. Because of their confined amount of data (3.8 MB), the single three-dimensional images can be easily transferred from one examiner to another on CD-ROM or by means of e-mail.6,28 Conventional radiographs and photographs allow an easier interpretation than threedimensional images because only two dimensions are considered at one time. However, the data gathered through conventional radiography or photography do not offer real measures because three-dimensional structures are projected onto a two-dimensional plane.8,17,29 As a consequence, a certain amount of information is inevitably lost, which may be crucial for the assessment of the adequate treatment concept. The evaluation of isolated lateral radiographs or photographs might supply sufficient information for treatment planning when a symmetric or nearly symmetric craniofacial situation is being projected. However, a great number of patients are operated on to achieve such ideal relationships.13,30,31 In the present study, different parasagittal and horizontal contours derived from the three-dimensional data were used to assess the symmetry of the midfacial advancement. These data could not have been acquired by the conventional methods. The measurements show that complete symmetry is difficult to achieve during the distraction period. The use of these contour measurements assessed after registra-
6 4 2 5 9 5.2 ⫾ 2.6 0.13
T2
6 1.5 2 3 5 3.5 ⫾ 1.9 0.10
T3
T3
0 0 2 0 0 0.4 ⫾ 0.9 0.10
Vertical
0 0 2 1 2 1.0 ⫾ 1.0 0.03
T2
Advancement Sn (mm)
8 4 3 6 5 5.2 ⫾ 1.9 0.01
T2
T3
T3
0 0 –1 0 0 –0.2 ⫾ 0.5 0.03
Vertical
0 0 –1 1 2 0.4 ⫾ 1.1 0.07
T2
Advancement LS (mm)
8 1.5 3 3 3 3.7 ⫾ 2.5 0.07
Sagittal T1
192 179 196 171 182 184.0 ⫾ 10.1 1.13
165 161 167 166 167 165.2 ⫾ 2.5 0.71
T2
N⬘SnPg⬘ (degrees)
T3
165 166 168 169 170 167.6 ⫾ 2.1 0.99
5.8 7.5 5.8 14 13 9.22 ⫾ 3.98 0.04
T2
T3
5.8 6 5.8 9 2 5.72 ⫾ 2.48 0.07
5.8 10.8 5.8 9 8 7.88 ⫾ 2.15 0.04
T2
Left T3
5.8 8 5.0 7 4 5.96 ⫾ 1.58 0.11
T2
0 4.7 4.4 3.8 5.0 3.58 ⫾ 2.05 0.06
T3
0 3 4.0 3 4 2.80 ⫾ 1.64 0.08
Right T2
0 6.8 4.4 6.1 5.6 4.58 ⫾ 2.71 0.04
Left
0 5 4.0 3 4 3.20 ⫾ 1.92 0.04
T3
Maximum Horizontal Advancement of Soft Tissue (mm)
T1, time 1 (postoperative); T2, time 2 (6 months after operation); T3, time 3 (2 years after operation).
1 2 3 4 5 Mean ⫾ SD Mean intraindividual SD
Patient
Right
Maximum Parasagittal Advancement of Soft Tissue (mm)
TABLE IX Soft-Tissue Advancement
2 2 1 3 4 2.4 ⫾ 1.1 0.04
T2
T3
2 2 3 1 2 2.0 ⫾ 0.7 0.06
Sagittal
1 1 1 1 2 1.2 ⫾ 0.5 0.08
T2
T3
1 1 1 1 1 1.0 ⫾ 0 0.04
Vertical
Advancement of Tip of the Nose (mm)
T1, time 1 (postoperative); T2, time 2 (6 months after operation); T3, time 3 (2 years after operation); Sn, subnasale point; LS, labrale superius; N⬘SnPg⬘, facial concavity (soft-tissue nasion to soft-tissue pogonion).
1 2 3 4 5 Mean ⫾ SD Mean intraindividual SD
Patient
Sagittal
TABLE VIII Soft-Tissue Advancement
Vol. 112, No. 2 / MIDFACIAL DISTRACTION
377
378
PLASTIC AND RECONSTRUCTIVE SURGERY,
August 2003
TABLE X Volume Data
Visible Volume Increase (cc) Right Patient
1 2 3 4 5 Mean ⫾ SD Mean intraindividual SD
Accommodation Vector (mm) Left
Right
Left
T2
T3
T2
T3
T2
T3
T2
T3
4.3 12.9 5.9 17.4 4.1 8.92 ⫾ 5.95 0.07
5.3 7.8 3.9 2.9 –2.2 3.54 ⫾ 3.70 0.07
4.3 13.7 7.4 14.5 7.8 9.54 ⫾ 4.39 0.04
5.3 9.3 6.7 1.6 1.1 4.80 ⫾ 3.47 0.04
2.95 4.9 3.4 7.6 3.2 4.41 ⫾ 1.94 0.06
2.6 3.6 2.6 0.6 –1.3 1.62 ⫾ 1.96 0.04
2.95 5.9 3.75 5.8 5.3 4.74 ⫾ 1.32 0.06
2.6 3.6 3.5 0.4 0.7 2.16 ⫾ 1.52 0.03
T2, time 2 (6 months after operation); T3, time 3 (2 years after operation).
tion of the three-dimensional images at different time points during the distraction period may constitute a promising approach toward further reduction of incongruencies between the left and right sides. The determination of volume changes in the malar-midfacial region completes the threedimensional information during the follow-up, when distraction osteogenesis is used. Despite the small number of patient cases, it seems that the hypothesis of distraction histiogenesis cannot be accepted easily. Although head circumference, body weight, and body height increased from time 1 to time 3 and the bony sagittal advancement represented by the incision superius showed a relapse of 19 percent between time 2 and time 3, the soft-tissue volume increase was reduced 50 percent in the same follow-up period (Tables III, VII, and X). In patient 5, there was even a loss of soft-tissue volume at time 3 compared with the preoperative situation. Scarring, muscle pull, soft-tissue tension, adaptation, and instability of bony fragments are the principal factors in relapse, especially in cleft lip and palate patients.32 The follow-up of large numbers of cases will help to address the question of distraction histiogenesis adequately, so that this hypothesis can be
accepted or rejected.3 If it should turn out that a clinically relevant soft-tissue relapse occurs frequently, overcorrection during the distraction period or an additional malar-midfacial augmentation should be added to the treatment concept. For further analysis of three-dimensional data, radial measurements have been proposed.4 These measurements do not follow the vector of advancement of the maxilla, which should be found in a midsagittal plane in an ideal case, but lie on a radial line drawn from the center of the skull. Because of the elliptic shape of the maxilla, it is easy to understand that the vector of advancement of the soft tissues away from the midline lies on neither a sagittal plane nor a radial line drawn from the center of the skull. Therefore, it seemed reasonable to the authors to calculate the accommodation vector that transfers the preoperative into the postoperative surface, and to correlate it with the bony movement of the maxilla. The accommodation vector takes advantage of every single data point in the region of interest and expresses the changes as a simple linear parameter. Therefore, the determination of the accommodation vector should be used as a standard measurement when three-
TABLE XI Soft-Tissue–to–Hard-Tissue Ratio
Ratio of Advancement Incision Superius to Labrale Superius Patient
1 2 3 4 5 Mean ⫾ SD Mean intraindividual SD
Ratio of Advancement Incision Superius to Vector Right
Ratio of Advancement Incision Superius to Vector Left
T2
T3
T2
T3
T2
T3
80 89 60 75 83 77.4 ⫾ 11.0 1.13
80 75 60 50 75 68.0 ⫾ 12.6 1.70
30 110 68 95 53 71.2 ⫾ 32.1 1.27
26 180 52 10 –33 47.0 ⫾ 80.5 0.70
30 130 75 73 88 79.2 ⫾ 35.8 0.85
26 180 70 7 18 60.2 ⫾ 71.1 0.85
T2, time 2 (6 months after operation); T3, time 3 (2 years after operation).
Vol. 112, No. 2 /
379
MIDFACIAL DISTRACTION
dimensional data have to be evaluated for facial analysis.
4.
CONCLUSIONS
It seems that besides classic linear and angular measurements, the determination of postsurgical changes of parasagittal and horizontal contours and soft-tissue volume is important for achieving a more complete evaluation of the facial shape. The contours can be selected individually to assess deviations from symmetry after bilateral distraction osteogenesis. For the measurement of volume changes and the accommodation vector, a well-defined region has to be chosen to collect reference data. It will be the aim of future research to define adequate norms for the three-dimensional changes of shape of the facial surface. At the moment, no reliable references are available. However, now that problems concerning measuring time, registration, and volume determination have been solved, the clinical application of the optical three-dimensional imaging technique can be extended to a large number of patients to establish reference databases. These data will be used for the accurate prediction of the longterm postoperative outcome after orthognathic surgery. Emeka Nkenke, M.D., D.M.D. Department of Oral and Maxillofacial Surgery University of Erlangen-Nuremberg Glueckstr. 11 91054 Erlangen, Germany
[email protected]
5.
6.
7.
8.
9.
10.
11.
12. 13.
ACKNOWLEDGMENTS This study was supported by the “Deutsche Forschungsgemeinschaft,” Special Research Sector 603, Model-Based Analysis and Visualization of Complex Scenes and Sensor Data, Subproject C4. The authors thank Dr. M. RadespielTro¨ ger of the Institute of Medical Informatics, Biometry, and Epidemiology, University of Erlangen-Nuremberg for his expert support during the statistical analysis.
14.
15.
REFERENCES 1. Hack, G. A., de Mol van Otterloo, J. J., and Nanda, R. Long-term stability and prediction of soft tissue changes after LeFort I surgery. Am. J. Orthod. Dentofac. Orthop. 104: 544, 1993. 2. Ko, E. W. C., Figueroa, A. A., and Polley, J. W. Soft tissue profile changes after maxillary advancement with distraction osteogenesis by use of a rigid external distraction device: A 1-year follow-up. J. Oral Maxillofac. Surg. 58: 959, 2000. 3. Fisher, E., Staffenberg, D. A., McCarthy, J. G., Miller, D. C., and Zeng, J. Histopathologic and biochemical changes in the muscles affected by distraction osteo-
16. 17.
18.
19.
genesis of the mandible. Plast. Reconstr. Surg. 99: 366, 1997. McCance, A. M., Moss, J. P., Fright, W. R., and Linney, A. D. Three-dimensional analysis technique: Part 3. Color-coded system for three-dimensional measurement of bone and ratio of soft tissue to bone: The analysis. Cleft Palate Craniofac. J. 34: 52, 1997. Aung, S. C., Ngim, R. C. K., and Lee, S. T. Evaluation of the laser scanner as a surface measuring tool and its accuracy compared with direct facial anthropometric measurements. Br. J. Plast. Surg. 48: 551, 1995. Bhatia, G., Vannier, M. W., Smith, K. E., Commean, P. K., Riolo, J., and Young, V. L. Quantification of facial surface change using a structured light scanner. Plast. Reconstr. Surg. 94: 768, 1994. Bush, K., and Antonyshyn, O. Three-dimensional facial anthropometry using a laser surface scanner: Validation of the technique. Plast. Reconstr. Surg. 98: 226, 1996. Ferrario, V. F., Sforza, C., Puleo, A., Poggio, C. E., and Schmitz, J. H. Three-dimensional facial morphometry and conventional cephalometrics: A correlation study. Int. J. Adult Orthod. Orthognath. Surg. 11: 329, 1996. Kawano, Y. Three dimensional analysis of the face in respect of zygomatic fractures and evaluation of the surgery with the aid of moire´ topography. J. Craniomaxillofac. Surg. 15: 68, 1987. Motegi, N., Tsutsumi, S., Okumura, H., Yokoe, Y., and Iizuka, T. Morphologic changes in the perioral soft tissues in patients with mandibular hyperplasia using a laser system for three-dimensional surface measurement. Int. J. Oral Maxillofac. Surg. 28: 15, 1999. Bohn, G., Cloutot, L., Habermeier, H. P., et al. Fast 3D-camera for industrial and medical applications. In Proceedings of ODIMAP II, 214 –219, 2nd Topical Meeting on Optoelectronic Distance/Displacement, Measurements and Applications, Pavia, Italy, May 20 –22, 1999. Häusler, G., and Gruber, M. Simple, robust and accurate phase measuring triangulation. Optik 89: 18, 1992. Keler, P., Wiltfang, J., Schultze-Mosgau, S., Hirschfelder, U., and Neukam, F. W. Distraction osteogenesis of the maxilla and midface using a subcutaneous device: Report of four cases. Br. J. Oral Maxillofac. Surg. 39: 13, 2001. Schultze-Mosgau, S., Schliephake, H., Schultze-Mosgau, S., and Neukam, F. W. Soft tissue profile changes after autogenous iliac crest onlay grafting for the extremely atrophic maxilla. J. Oral Maxillofac. Surg. 58: 971, 2000. Falkner, F. Growth monitoring: Fetus to the first two postnatal years. In O. Brusner, F. R. Carrazza, M. Gracey, B. L. Nichols, and J. Senterre (Eds.), Clinical Nutrition of the Young Child. New York: Raven, 1985. Pp. 28 –38. Houston, W. J. B. The analysis of errors in orthodontic measurements. Am. J. Orthod. 83: 382, 1983. Gaggl, A., Schultes, G., and Kärcher, H. Changes in soft tissue profile after sagittal ramus osteotomy and repositioning of the mandible. J. Oral Maxillofac. Surg. 57: 542, 1999. Kakoschke, D., Gäbel, D., and Schettler, D. Three-dimensional photogrammetric acquisition of facial surfaces. Mund. Kiefer. Gesichtschir. 1: 61, 1997. Moss, J. P., McCance, A. M., Fright, W. R., Linney, A. D., and James, D. R. A three-dimensional soft tissue
380
20.
21.
22.
23.
24.
25.
analysis of fifteen patients with class II, division 1 malocclusions after bimaxillary surgery. Am. J. Orthod. Dentofac. Orthop. 105: 43, 1994. Girod, S., Keeve, E., and Girod, B. Advances in interactive craniofacial surgery planning by 3D simulation and visualization. Int. J. Oral Maxillofac. Surg. 24: 120, 1995. Girod, S., Keeve, E., and Girod, B. Soft tissue prediction in orthognathic surgery by 3D CT and 3D laser scanning. In Proceedings of the XIIth Congress of the European Association for Craniofacial Surgery, Den Haag, Netherlands, May 5–7, 1994. Ewing, M., and Ross, R. B. Soft tissue response to orthognathic surgery in persons with unilateral cleft lip and palate. Cleft Palate Craniofac. J. 30: 320, 1993. Freihofer, H. P. M. The lip profile after correction of retromaxillism in cleft and non-cleft patients. J. Maxillofac. Surg. 4: 136, 1976. Genecov, R. G., Sinclair, P. M., and Dechow, P. C. Development of nose and soft tissue profile. Angle Orthod. 60: 191, 1990. Behrents, R. G. Growth in the Aging Craniofacial Skeleton: Craniofacial Growth Series, Monograph 17. Ann Arbor,
PLASTIC AND RECONSTRUCTIVE SURGERY,
26.
27. 28.
29. 30.
31.
32.
August 2003
Mich.: Ann Arbor Center for Human Growth and Development, University of Michigan, 1985. McCance, A. M., Moss, P. J., Wright, W. R., Linney, A. D., and James, D. R. A three-dimensional soft tissue analysis of 16 skeletal class III patients following bimaxillary surgery. Br. J. Oral Maxillofac. Surg. 30: 221, 1992. Mardia, K. V. Comment: Some contributions to shape analysis. Stat. Sci. 4: 87, 1989. Nkenke, E., Cloutot, L., Laboureux, X., Häusler, G., and Neukam, F. W. 3D-scanning of facial surfaces for determination of volume changes in soft tissue. J. Craniomaxillofac. Surg. 26: 133, 1998. Cohen, A. M. Uncertainty in cephalometrics. Br. J. Orthod. 11: 44, 1984. Cedars, M. G., Linck, D. L., Chin, M., and Toth, B. A. Advancement of the midface using distraction techniques. Plast. Reconstr. Surg. 103: 429, 1999. Molina, F., and Ortiz-Monasterio, F. Mandibular elongation and remodeling by distraction: A farewell to major osteotomies. Plast. Reconstr. Surg. 96: 825, 1995. Hochban, W., Ganss, C., and Austermann, K. H. Longterm results after maxillary advancement in patients with clefts. Cleft Palate Craniofac. J. 30: 237, 1993.
