| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Scientific Section |
Department of Orthodontics, Dental Institute, Royal London Hospital, Whitechapel, London, E1 2AD, U.K.
Abstract
The objective of the investigation was to study the three-dimensional growth of the face, and to examine the hypothesis that there are three-dimensional differences between the faces of boys and girls. The subjects comprised 132 British Caucasians aged 5–10 years measured by optical surface scanning in this cross-sectional study. Average scans for each age and sex subgroup were superimposed to assess the differences with age and sex. Males were generally larger than females. The greatest difference was between the facial heights and the least in the mid-facial dimensions. The face height of both sexes increased by an average of 3–4 mm annually. Mid-face prominence and width altered little. Mandibular width increased by 1–3 mm a year, rising to 3–5 mm in some years at the inferior areas of the mandibular region. Mandibular prominence also increased. Nose height and prominence and alar base width increased by 2 mm per year on average. Dorsum width changed little. Boys were generally larger than girls. Growth in facial height was greatest. Mid-face prominence and width changed little with age, whilst the prominence and width of the lower face increased more. Nasal prominence and alar base width increased at most ages. Dimensions changed more than reported by cephalometric studies, possibly as this study included the soft tissues.
Refereed Scientific Paper
Key words: Facial Growth, Childhood, Optical Surface Scanner, Three-dimensions
Introduction
The volume of literature reporting research into facial growth is huge, reflecting the importance of the topic. Much of the measurement has been two-dimensional, but some studies have measured in three dimensions.
Vertical skeletal growth has been measured by cephalometrics and anthropometrics, and average annual changes of 1–2 mm in the pre-pubertal period have been found (Riolo et al., 1974
; Farkas, 1981; Van der Beek et al. 1991). Using photography, Bishara et al. (1995) showed an increase of just under 2 mm a year. However, the method of landmark identification by eye on photographs could be a potential source of error. The sex difference in facial height varies from 1 mm (Nanda, 1955; Snodell et al., 1993) to just over 2 mm, with males larger than the females (Riolo et al., 1974; Bhatia and Leighton, 1993). Bishara et al. (1995) found the male face height to be 5–6 mm greater than the female. This could reflect the inclusion of the soft tissues in their measurements or the difficulty of accurately placing landmarks on a two-dimensional photograph.
Less data has been collected on facial width, due to the predominance of lateral cephalometry as an imaging technique. Woods (1950) found an average annual bi-gonial width increase of 1•5 to 2 mm. Farkas (1981) found the zygomatic and gonial widths each increased by about 7 mm from 6 to 10 years of age. As the gonial width was initially less, it proportionally changed more. This agrees with Snodell et al. (1993), who found a gradient of width growth, with the cranium growing least and the mandible the most. A number of studies have shown width increases to be less than vertical changes (Snodell et al., 1993; Bishara et al., 1995). Sex differences in zygomatic width of about 1 mm (Farkas, 1981) and bi-gonial width of 3 mm (Riolo et al., 1974; Farkas, 1981) have been found.
The soft tissue facial contour has been shown to change differently from the underlying hard tissues (Subtelny, 1959). The soft tissue convexity increases slightly from 3 to 6 years and remains quite stable after that, despite decreasing hard tissue convexity. This is due to the soft tissues overlying A point growing in thickness more than those overlying pogonion. This effect excludes the nose, whose forward growth causes an increasing convexity after 3 years of age. Chaconas and Bartroff (1975) also found that the increase in facial convexity was largely due to nasal growth. Woods (1950) found an increasing gradient of facial growth in terms of prominence, with the sella–nasion line attaining 95 per cent of adult length by five years, the maxilla 87 per cent and the mandible 81 per cent. The soft tissue integument of males is thicker than females at all pre-pubertal ages (Subtelny, 1959).
Many papers have reported on nasal growth, finding it to be an important contributor to changes in the overall facial profile. An increase in nasal size has been reported by many researchers (Subtelny, 1959; Burke and Hughes-Lawson, 1989; Nanda et al., 1990), although lateral cephalometry has been constrained to vertical and horizontal changes. Burke and Hughes-Lawson (1989), using stereophotogrammetry, and Snodell et al. (1993), using postero-anterior cephalometry, both found that nasal width grows markedly. Farkas (1981), and Burke and Hughes-Lawson (1989) both showed little change in dorsum width and a greater change in alar width, the latter concluding that the inter-canthal width relates to orbital growth. This would cause it to follow a neural growth pattern, and thus be nearer to its adult size than other parts of the nose. Burke and Hughes-Lawson (1989) showed a similar volumetric change in the sexes until 9 years of age, after which the male nose grew at a greater rate. The sex difference in nasal dimensions has generally been found to be quite small in the pre-pubertal period (Farkas, 1981; Nanda et al., 1990; Snodell et al., 1993), the male nose being up to 1 mm larger in most of its dimensions.
Much of the research on facial growth has been performed using lateral cephalometry (Riolo et al., 1974; Bhatia and Leighton, 1993). Whilst it is an accurate reproducible method, there are shortcomings. The radiograph is a two-dimensional representation of a three-dimensional subject, leading to a simplified interpretation of the structures imaged and the process of facial growth. There is the inherent problem of measuring growth of three-dimensional curved surfaces by linear and angular measure-ments between artificially defined landmarks (Moyers and Bookstein, 1979). Structures out of the mid-saggital plane are subject to increased errors of magnification and identification (Steiner, 1953).
The cephalostat was originally devised to take postero-anterior, as well as lateral views (Broadbent, 1931), giving the ability for three-dimensional cephalometric measurements (Grayson et al., 1988). However, most soft tissue points are not available for study as they do not appear on both views. Additionally, geometrically determined landmarks, such as menton, will be positioned in different locations depending on the direction from which the radiograph is taken. All the landmarks have different magnification factors in each view, as they are a different distance from the source and the film. Stereophotogrammetric cephalometry, using radiographs from two different views, has been developed as a method of three-dimensional analysis (Baumrind et al., 1983), but the technique can be difficult and time-consuming to use.
Other methods of three-dimensional measurement of the face have been employed. Anthropometry (Farkas, 1981) requires only simple instruments and is an easy technique to learn. However, it is time-consuming to carry out comprehensive measurements, making it less suitable for the young subject. Obviously, it does not produce an image of the face, only linear and angular measurements. Photographic techniques suffer from image distortion and inaccuracy of measurement unless a telecentric lens is used (Robertson and Volp, 1981). Techniques relying on the digitization of contour maps such as moiré stripe photography (Takasaki, 1970; Kawai et al., 1990) can be difficult to analyse and measure, especially in areas of high relief such as the nose, where the contour lines will be very closely placed. Measurement by stereophotogrammetry (Burke and Beard, 1967) has been applied to the face, but the technique is difficult and requires considerable expertise.
The optical surface scanner (Arridge et al., 1985) has the advantage of rapid accurate data collection in three dimensions. The lack of ionising radiation allows data to be acquired of normal subjects for research. The associated software creates a visual image of the subject and an average image for a group of subjects, thus allowing differences between subjects or groups of subjects to be measured. The accuracy of the system has been tested (Aung et al., 1995). The major disadvantage of the system is its cost.
Most recently, a video-capture stereoscopic method of imaging has been developed (Ayoub et al., 1998). Two pairs of stereo cameras ensure that the curved facial surfaces are completely imaged within 2 seconds. The system allows photo-realistic image generation of the face that can be viewed from any direction. The polygonal facial model can be used to measure facial landmarks and volumes. An accuracy of landmark capture of 0•5 mm has been quoted and the system promises much for the future in terms of facial imaging, measurement, and surgical prediction.
The aims of this study were to compare the three-dimensional facial morphology of girls and boys of the same age, and to study the three-dimensional alterations in the facial morphology with increasing age.
Subjects and Methods
Data acquisition
One-hundred-and-thirty-two 5–10-year-olds comprising 72 males and 60 females, were recruited for the study. They were distributed as shown in Table 1.
|
It was felt that without these, a child with abnormal facial morphology could have biased the average for their age and sex subgroup.
Each child was scanned once by the optical surface scanner (Figure 1). The subject was instructed to sit still with the facial muscles relaxed and the eyes lightly closed. With the room darkened, a lens fanned the helium-neon laser beam into a 0•7-mm wide line along the face. As the motorized chair rotated, the face was scanned by the laser over a period of about 10 seconds. A video camera read approximately 60,000 data points from across the face. A microcomputer processed the data points and joined adjacent points to form triangular facets. These facets were shaded to give a solid-looking image that could be viewed on a monitor. If the image obtained was of poor quality, the scan was repeated until one of a sufficiently high standard had been obtained.
|
Five further points were automatically constructed by the computer across the forehead. The software program joined landmarks 1 to 5 by a line of best fit. A second, parallel line was then constructed thirty mm above this. A landmark was constructed on this line, above landmark 3. A further landmark was constructed 15 mm either side of this and a further landmark 15 mm lateral to each of these.
The final 5 points, like the first 5 were marked by the operator:
Averaging
Each scan was edited by the operator to remove unwanted data, such as the hair and the base of the neck, as during the averaging procedure the computer treats all areas of the scan image with equal importance.
The 15 landmarks were used by the computer to create a representative average scan for each age and sex subgroup. The set of scans were resampled on a cylindrical co-ordinate grid system, and the co-ordinates averaged to give the average scan and a one standard deviation facial form. The average for the 8-year-old males is shown from the frontal and right lateral views as an example (Figure 2).
|
Each pair of average scans being superimposed and compared was displayed adjacently on a monitor. ‘Warm’ colours (yellow
orangered) indicated positive differences and ‘cold’ colours (greenbluepurple) indicated negative differences. Each colour represented a 2-mm difference in dimensions from the previous colour. Areas with no dimensional difference were shaded brown. The left side of the screen displayed the result of the left image compared to the right, and the right side of the screen displayed the result of the right image compared to the left. Thus, the colour shading of one image was the exact opposite of the other. The superimposition images were produced in both the frontal and the lateral views.
Averages of the two sexes at each age were superimposed to examine the differences between the sexes. Averages of the same sex, but different ages were superimposed to examine the changes with increasing age.
Error Study
This was performed to assess the error due to scan acquisition and landmarking. Ten adults were scanned twice, with a 1-week interval between the scans. Each group of 10 scans was averaged and the two averages were superimposed. The superimposition was mostly brown, denoting no dimensional difference and, hence, indicating a consistent technique of scan acquisition and landmarking (Figure 3).
|
The superimposition of the seven year old male and female averages illustrates how the results were analysed (Figure 4). The colour bar at the bottom of the picture shows the dimensional difference represented by each colour. The male face height is 7–9 mm greater than the female. There is very little difference in the infra-orbital regions in either width or prominence. The male face is 1–3 mm wider generally, increasing to 3–5 mm at the inferior region of the mandible and 5–7 mm at the lower border. The lower half of the face and much of the chin are 1–3 mm more prominent in the males than the females. The male alar base is 1–3 mm wider than the female on the right side, but less on the left side. The height and prominence of the noses are similar.
|
Female Changes with Age
The superimposition of the 9- and 10-year-old averages is shown in Figure 5 as an example. The face height increased each year. The actual amount varied from year to year, but averaged approximately 4 mm each year. The width and prominence of the mid-face varied very little each year. The width of the lower parts of the mid-face and the mandible increased more, with a 1–3-mm increase at most ages, and 3–5 mm in some of the most inferior parts of the mandibular region. The prominence of the chin varied little until eight years of age. After this age, the chin increased in prominence: by 3–5 mm from 9 to 10 years of age. The nose increased by 1–3 mm in most of its dimensions each year.
|
as an example. The pattern of facial growth was very similar to that of the females. The face height increased by a variable annual amount, which averaged approximately 4 mm. The width of the mid-face varied little, only increasing by 1–3 mm from 5 to 6 and 7 to 8 years. The width increase of the lower part of the face was greater, rising to 3–5 mm in the most inferior regions. The prominence of the mid-face varied little, whilst the chin became 5–7 mm more prominent from 5 to 10 years. As with the females, the majority of the increase in prominence occurred after 8 years of age. The nose increased in width by 1–3 mm annually. The increase in nose height was generally greater, rising to 3–5 mm from 9 to 10 years of age.
|
There was very little difference between the nasal dimensions in the 5-, 6-, and 7-year age groups. However, in the older age groups the alar base width is 1–3 mm greater in the males. The 10-year-old male nose was 1–3 mm greater in all the dimensions of height, width, and prominence.
Discussion
This study was cross-sectional in design. The subjects used can, however, be re-scanned each year and so this data serves as the basis for a longitudinal study. It is quite possible that a child who was large or small for their age could have affected the average for their subgroup, thus giving rise to misleading results. This could explain some of the fluctuations in the annual changes for some measurements. Similarly, the 9-year-old boys were found to be smaller than the 9-year-old girls and the 8-year-old boys, in direct opposition to the pattern found for all other age and sex groups. This is most likely to be due to some small subjects in the 9-year-old male subgroup.
A subgroup containing a higher proportion of subjects at either the upper or lower end of the year's range could make the average larger or smaller, respectively, than the true average for that age. To minimize this problem, as many children of each age were scanned as possible and the average age of the subgroups was generally close to the middle of each year's range (Table 1
).
All the landmarking was performed by one operator for consistency. The visualization of a landmark in three dimensions is not possible from a two-dimensional image. Therefore, the raw data points that make up the vertical and transverse profiles for the area to be landmarked are also viewed. This allows the location of points of maxima and minima on the curving surfaces of the face, and thus allows three-dimensional landmarks to be defined. As with any technique relying upon the use of landmarks, there is the possibility of operator error. In recognition of this fact, methods of mathematically defining the landmarks are being developed, which should hopefully lead to the automation of this procedure by the computer software (Moss et al. 1992).
Changes with Age
The increase in face height in both sexes is greater than found by Farkas (1981), Riolo et al. (1974), and Van der Beek et al. (1991), who found an annual change of approximately 2 mm. The difference could be due to this study including the soft tissues, whereas the other studies measured only the hard tissue face height.
The small change in mid-face width agrees with other studies (Farkas, 1981; Bishara et al., 1995). The greater increase in the mandibular width also agrees with other studies (Riolo et al., 1974; Farkas, 1981; Bishara et al., 1995), but the magnitude of the changes found was greater in this study. As mentioned before, this could be because the cephalometric and anthropometric studies aim to measure only the hard tissues. Most studies measure the bi-gonial width and, indeed, this is the only measure of mandibular width studied by cephalometry. In contrast, this study included the soft tissues such as the masseter, which will also increase in dimensions with age, in the same way as the hard tissues. One of the advantages of the three-dimensional collection of data performed by the optical surface scanner was its ability to provide a more complete view of the width changes of the whole face than is available by cephalometry.
A gradient of change in facial width can therefore be seen. There is little or no increase in the upper face, small changes in the mid-face and, increasingly, greater increases in the lower face as the lower border of the mandible is approached.
Comparison of changes in facial prominence with other studies is not possible, as cephalometry usually measures some type of maxillary or mandibular unit length. However, the pattern of change, with small increases in the mid-face and greater increases in the lower face, agrees with other studies (Canon, 1970; Nanda, 1992).
The increase in alar base width, nasal height, and nasal prominence, but no increase in the dorsum width is in agreement with other studies (Meng et al., 1988; Burke and Hughes-Lawson, 1989). However, the annual alar base width increase of 1 mm bilaterally is more than Farkas (1981) and Bishara et al. (1995) who reported an increase of 0•5 mm per year. This may be due to the scan time allowing alar flaring with breathing. The increases in nasal height and prominence agree with other studies.
Comparison of Males and Females
The greater male face height agrees with other studies. However, Farkas (1981) found only 2 mm difference in the total face height using anthropometry. This, however, was with the soft tissues compressed over gnathion. Cephalometric studies also report less difference, possibly as they only measure the hard tissue face height. Bishara et al. (1995) used photography and found a more comparable difference of 5–6 mm. Whilst the soft tissues may be responsible for some of the difference between this study and others, it is considerable. This is difficult to account for, but the difference was consistent throughout the ages.
The small difference in the mid-facial widths is in agreement with other studies. Again, this study shows slightly greater differences than some, possibly as the soft tissues were included. The optical surface scanner measurement of width difference in the gonial region is generally greater than the cephalometric or anthropometric studies. As discussed before, this is likely to be as the soft tissues, such as the masseter, are included in the optical surface scanner measurements, but not the other studies.
As discussed earlier, other studies do not give comparable data on facial depth, as cephalometry usually measures some type of maxillary or mandibular length. Prominence of the cheek regions cannot be measured with cephalometry.
The small sex difference in the nasal dimensions agrees with the stereophotogrammetric study by Burke and Hughes-Lawson (1989), and the photogrammetric study by Bishara et al. (1995).
Conclusions
Acknowledgments
I would like to thank Miss P. Goodwin and the staff of the department of Medical Physics at University College London for their teaching in the operation of the optical surface scanner and its software.
References
Arridge, S., Moss, J. P., Linney, A. D. and James, D. R. (1985) Three dimensional digitization of the face and skull, Journal of Maxillofacial Surgery, 13, 136–143.[Medline]
Aung, S. C., Ngim, R. C. K. and Lee, S. T. (1995) Evaluation of the laser scanner as a surface measuring tool and its accuracy compared with direct facial anthropometric measurement, British Journal of Plastic Surgery, 48, 551–558.[Medline]
Ayoub, A. F., Siebert, P., Moos, K. F., Wray, D., Urquart, C. and Niblett, T. B. (1998) A vision-based three-dimensional capture system for maxillofacial assessment and surgical planning, British Journal of Maxillofacial Surgery, 36, 353–357.
Baumrind, S., Moffit, F, H. and Curry, S. (1983)Three-dimensional X-ray stereometry from paired coplanar images: a progress report,American Journal of Orthodontics, 84, 292–312.[Medline]
Bhatia, S. N. and Leighton, B. C. (1993)A Manual of Facial Growth. A Computer Analysis of Longitudinal Cephalometric Growth Data,Oxford University Press, Oxford.
Bishara, S. E., Jorgensen, G. J. and Jakobsen, J. R. (1995) Changes in facial dimensions assessed from lateral and frontal photographs. Part II—Results and conclusions, American Journal of Orthodontics and Dentofacial Orthopedics, 108, 489–499.[Medline]
Broadbent, B. H. (1931) A new x-ray technique and its application to orthodontia, Angle Orthodontist, 1, 45–66.
Burke, P. H. and Beard, L. F. H. (1967) Stereophotogrammetry of the face, American Journal of Orthodontics, 53, 769–782.[Medline]
Burke, P. H. and Hughes-Lawson, C. A. (1989) Stereophotogrammetric study of growth and development of the nose, American Journal of Orthodontics and Dentofacial Orthopedics, 96, 144–151.[Medline]
Canon, J. (1970) Craniofacial height and depth increments in normal children, Angle Orthodontist, 40, 202–218.
Chaconas, S. J. and Bartroff, J. D. (1975) Prediction of normal soft tissue changes, Angle Orthodontist, 45, 12–25.[Medline]
Farkas, L. G. (1981)Anthropometry of the Head and Face in Medicine,Elservier, New York.
Grayson, B., Cutting, C., Bookstein, F. L., Kim, K. and McCarthy, J. G. (1988) The three-dimensional cephalogram: theory, technique, and clinical application, American Journal of Orthodontics and Dentofacial Orthopedics, 94, 327–337.[Medline]
Kawai, T., Natsume, N., Shibata, H. and Yamamoto, T. (1990) Three dimensional analysis of facial morphology using Moiré stripes. Part I. Method, International Journal of Oral and Maxillofacial Surgery, 19, 356–358.[Medline]
Meng, H. P., Goorhuis, J., Kapila, S. and Nanda, R. S. (1988) Growth changes in the nasal profile from 7 to 18 years of age, American Journal of Orthodontics and Dentofacial Orthopedics, 94, 317–326.[Medline]
Moss, J. P., Campos, J. C. and Linney, A. D. (1992)The analysis of profiles using curvature analysis,European Journal of Orthodontics, 14, 457–461.
Moyers, R. E. and Bookstein, F. L. (1979) The inappropriateness of conventional cephalometrics, American Journal of Orthodontics, 75, 599–617.[Medline]
Nanda, R. S. (1955) The rates of growth of several facial components measured from serial cephalometric roentgenograms, American Journal of Orthodontics, 41, 658–673.
Nanda, S. K. (1992) Differential growth of the female face in the anteroposterior dimension, Angle Orthodontist, 62, 23–34.[Medline]
Nanda, R. S, Meng, H., Kapila, S. and Goorhuis, J. (1990) Growth changes in the soft tissue facial profile, Angle Orthodontist, 60, 177–190.[Medline]
Riolo, M. L., Moyers, R. E., MacNamara, J. A. and Hunter, W. S. (1974)An Atlas of Craniofacial Growth: cephalometric standards from the University of Michigan growth study, Monograph 2, Craniofacial Growth Series, Ann Arbour: Centre for Human Growth and development, University of Michigan.
Robertson, N. R. E. and Volp, C. R. (1981) Telecentric Photogrammetry: Its development, testing and application, American Journal of Orthodontics, 80, 623–637.[Medline]
Snodell, S. F., Nanda, R. S. and Currier, G. F. (1993) A longitudinal cephalometric study of transverse and vertical craniofacial growth, American Journal of Orthodontics and Dentofacial Orthopedics, 104, 471–483.[Medline]
Steiner, C. C. (1953) Cephalometrics for you and me, American Journal of Orthodontics, 39, 729–755.
Subtelny, J. D. (1959) A longitudinal study of soft tissue facial structures and their profile characteristics, defined in relation to underlying structures, American Journal of Orthodontics, 45, 481–507.
Takasaki, H. (1970) Moiré Topography, Applied Optics, 9, 1461–1472.
Van der Beek, M. C. J., Hoeksma, J. B. and Prahl-Andersen, B. (1991) Vertical facial growth: a longitudinal study from 7 to 14 years of age, European Journal of Orthodontics, 13, 202–208.
Woods, G. A. (1950) Changes in width dimension between certain teeth and facial points during human growth, American Journal of Orthodontics, 36, 676–699.
This article has been cited by other articles:
|
|
 |
|
  B. K. Cha, J. Y. Lee, P.-G. Jost-Brinkmann, and N. Yoshida Analysis of tooth movement in extraction cases using three-dimensional reverse engineering technology Eur J Orthod, August 1, 2007; 29(4): 325 - 331. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Moss The use of three-dimensional imaging in orthodontics Eur J Orthod, October 1, 2006; 28(5): 416 - 425. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
