N. V. Hermann, DDS, PhD (1, 2); T. A. Darvann, MSc (3); R. Fritz, DDS (4);  R. Long JR., DMD, PhD (4);
J. L. Marsh, MD (5); D. V. Huebener, DDS, MS (6); S. Kreiborg, DDS, PhD, Dr Odont (1, 7)

(1)   Department of Pediatric Dentistry, School of Dentistry, University of Copenhagen, Copenhagen, Denmark
(2)   Department of Oral Function and Physiology
(3) 3D-Laboratory, Department of Informatics and Mathematical Modelling, Technical University of Denmark,
Copenhagen University Hospital, and School of Dentistry, University of Copenhagen, Copenhagen, Denmark
(4) Lancaster Cleft Palate Clinic, Lancaster, PA, USA
(5) Department of Pediatric Plastic Surgery, St. Louis Children's Hospital, St. Louis, MO, USA
(6) Department of Pediatric Dentistry, St. Louis Children's Hospital, St. Louis, MO, USA
(7) Department of Pediatrics 1, Juliane Marie Center, Copenhagen University Hospital, Copenhagen, Denmark
 
 

ABSTRACT

The current paper reports on a statistical comparison of maxillary arch form as measured on plaster casts from two
groups of 9-year-old UCLP children:  20 individuals treated with infant orthopedics, and 20 individuals treated without.
The particular method utilized for the measurement of arch form, including data acquisition and subsequent analysis, is,
as a whole, referred to as the method of 3D CT scan analysis. The major advantageous features of this method, in
addition to fast data acquisition, are accurate 3D measurements and the ability to describe the arch form in individuals
with teeth.  The current paper demonstrates the method's ability to estimate mean arch form.  The statistical comparison
of the mean arch forms in the two groups (each of relatively small sample sizes; n = 20) was carried out using selected
simple variables (distances, angles, ratios) derived from the computed arch forms. Only two variables revealed a
significant difference at a 5% significance level using Student's t-test.  However, several variables were close to being
significant, indicating that at least some of the observed differences between arch forms could be real. A conclusive
result would depend on either an increase of sample size or, possibly, a more elaborate statistical analysis.
 

INTRODUCTION

Cleft lip and/or palate are the most frequent congenital malformations in the craniofacial complex (1:500 births),
although inter-racial differences in incidence occur (Gorlin et al. 1990). Approximately a quarter million new babies
with clefts are born each year (Lee 1999).

Infant orthopedics (McNeil 1956) has existed for more than 50 years and is widely used in many different ways and
forms; varying in duration and combination with various surgical treatment regimes (Friede and Katsaros 1998). Infant
orthopedics is, however, still very controversial (Winters and Hurwitz 1995; Berkowitz 1996) due to the difficulties in
quantifying the claimed effects, which are: a) narrowing the cleft and molding the segments into a more symmetric
configuration before primary lip surgery (presurgical orthopedics) in order to facilitate the surgical intervention, b)
prevention of collapse of the segments after primary surgery (postsurgical), c) normalization of facial growth, d)
improvement of feeding habits, e) psychological support for the parents, and f) better speech development.  In general,
studies of infant orthopedics have shown great variability in the treatment outcome, varying from very good results to
no measureable effect. The major reasons for this could be: a) Sampling problems, especially related to sex, age, ethnic
background, inclusion of "syndromic" cleft cases, cleft subtypes, different surgeons, sample size, control data. b)
Methodological problems; e.g. different ways of measuring the maxillary changes, the type of infant orthopedic
appliances and changing surgical treatment regimes.

The computer technology has facilitated the development of methods to quantify plaster casts of cleft palate cases in
three dimensions. The usefulness of some of these methods are limited because each landmark  has to be manually
digitized on the model. Examples of digital type probes are the Reflex Microscope (Kriens 1991; Ball et al. 1995) the
Perceptor (Berkowitz 1990); and the Optocom (Van der Linden et al. 1972; Kuijpers-Jagtman 1983, 1989). Recently
developed three-dimensional methods have overcome some of these problems (Kilpelainen and Laine-Alava 1996;
Bacher et al. 1998; Ferrario et al. 1998). Also, the introduction of the surface laser scanning of plaster casts (e.g. Foong
et al. 1999; Stellzig et al. 1999) has made it possible to capture an accurate three dimensional model of a plaster cast as
well as to carry out accurate three-dimensional measurements and storage of models over time.  However, the
acquisition time is long and only one model can be scanned at a time (Cyberware 1999).

The purpose of the present study was to evaluate effects of two infant management protocols for unilateral complete
cleft lip and palate (UCLP) on maxillary arch form and dimensions at mixed dentition using a new method involving
computer recontructed models of the maxilla from CT scanned plaster casts (Hermann et al. 1999).
 

MATERIAL AND METHODS

Sample

Two longitudinal samples of children with unilateral complete cleft lip and palate (UCLP) treated with two different
infant management protocols were included in the study.

Except from the criterion of diagnosis, the criteria for inclusion in the study groups were as follows:
1)  The infant was of Unites States of America Caucasian origin.
2)  The child was healthy except for its single cleft malformation.
3)  Institution A children were treated with definitive lip and then palate repairs, only.
4)  Institution B children were treated with an initial lip adhesion + passive molding plate,  followed by subsequent
definitive lip and then palate repairs.
5)  The children from Institution A and B were matched according to cleft type, gender (11 males and 9 females were
included in each group), and initial alveolar cleft gap (Institution A: 8.75 mm +/- 2.46; Intitution B: 8.48 mm +/- 2.27).
6) The age of examination for all children included in the study was about 9 years (Institution A: 9.00 years +/- 1.30;
Institution B: 9.19 years +/- 1.47), just prior to orthodontic intervention in mixed dentition.

The sample consisted of a total number of 40 subjects; 20 UCLP subjects from Institution A, and 20 UCLP subjects
from Institution B.
 
 

DATA ACQUISITION, LANDMARK DIGITIZATION, AND VARIABLE DEFINITION

Impression of the maxilla was obtained just prior to the orthodontic intervention and maxillary plaster casts were
produced. The plaster casts were duplicated identically, and randomly numbered for blinding.

The maxillary plaster casts were CT-scanned in a SIEMENS spiral CT-scanner; 9 plaster casts were scanned at the
same time with a slice thickness of 1 mm and a slice distance of 0.5 mm.  Slice images were reconstructed at every 0.5
mm, and a 3D volume image was created.  Iso-intensity reconstructions corresponding to the surface of the plaster casts
were created in the software Mvox (TM).  Thirty-nine landmarks were digitized interactively on the surface
reconstructions according to a previously defined sequence in 3D by use of Mvox (TM). Nineteen variables (13 linear,
3 angular, 3 ratios) were computed by use of IDL (Interactive Data Language) (IDL 1997). The variable definitions are
given in Table 1.  Some variables were directly based on the digitized landmarks, others were derived from several
landmarks.  In particular, a number of landmarks along each alveolar arch segment were used in order to compute a 3D
spline curve representing the form of the segments. The alveolar arch, as represented by the fitted spline curves and
selected landmarks, was visualized separately, as well as together with the surface representation of the plaster cast in
order to inspect the morphology of the cast, see Figure 1.

Figure 1. Transparent computer model of a plaster cast with 3D spline curve fitted to the digitized  landmarks.
 
 
 

STATISTICAL CALCULATIONS

Right clefts were mirrored in order to have the cleft on the left side.  Figure 2 shows the computed spline curves
representing the maxillary alveolar arch in the individuals in the two groups. Superimposition of the curves was carried
out using the midpoint of the line connecting the Sillman (tuber) points (SID and SIS), representing the x-axis, and the
point M where a plane perpendicular to the SID-SIS line intersects the spline curve of the large segment.  The three
points SID, SIS and M define the x-y plane. No relative scaling of the curves was carried out. Mean alveolar arches
were calculated, represented by mean alveolar spline curves.  Mean values of the 19 variables in each of the two groups
were compared by a two-tailed Student's t-test, as well as by a Wilcoxon Rank-Sum test, and variances were compared
by F-test.  The level of significance was set to 5%.

MOVIE

Figure 2. The spline curves for all models included in the two groups (Institution A and B).
 
 

RESULTS
 

Figure 3 shows two views of the mean alveolar arches.    Mean values for all the computed variables are shown in
Table 1.  Distances between the curves representing the Institution A and Institution B samples, respectively, are
plotted in Figure 4.  In the figure the distances between curves are plotted as a function of the position along the arch,
and separately for x-, y-,  z- and a resultant 3D-distance.  The figure shows that the maximum distance between the two
arches is about 1.3, 2.0 and 0.9 mm in the x-, y-, and z-directions, respectively, and the maximum 3D distance is about
2.1 mm.  p-values for the statistical comparison of the mean values of the variables are also presented in Table 1.
Mean values of two of the variables were significantly different: 1) ASID; the angle between the vector SID-ALVD
and the vector SID-SIS, indicating that the lateral segment on the cleft side was rotated more medially in Institution B
than in Institution A casts; 2) RWL; the ratio between the posterior width (SID-SIS) and the maxillary length (LM),
indicating that this ratio is largest in Institution B casts.  In addition to these significant findings, visual inspection of
the mean arch shapes, together with observing that several variables showed near-significant differences, indicated a
trend towards Institution B casts being slightly wider posteriorly and somewhat more narrow anteriorly than in
Institution A. Furthermore, that Institution A arches had greater length in the sagittal dimension.   Visual inspection of
Figure 2 could indicate that the arches of Institution A has a larger variability than Institution B arches; however, this
was not supported by the F-test (Table 1).
 

MOVIE
 

Figure 3. The mean alveolar spline curves for the two groups shown in two different orientations.
 
 
 
 

Table 1

Name	Description	Mean (A)	Mean (B)	SD (A)	SD (B)	p_paired(t) / p(t) / p(r)
LS	Length of Small Segment	40.89	39.59	3.89	3.22	0.27 /  0.26  / 0.15
LL	Length of Large Segment	64.38	61.40	5.59	4.73	0.07 / 0.08 / 0.07
DS	Distance between Sillmann points	42.19	44.65	5.77	4.30	0.13 / 0.14 / 0.12
DA	Distance between Alveolar points	1.24	1.18	1.73	1.16	0.90 / 0.90 / 0.57
LM	Length of Maxillary Midline	41.03	39.21	3.84	3.49	0.15 / 0.12 / 0.09
S%	Small Segment: % Coverage of LM	87.24	87.41	4.76	4.82	0. 92 / 0.91 / 0.83
WM	Max Width of Maxilla	43.49	45.35	5.04	3.73	0.16 / 0.19 / 0.19
WM%	Position of Max Width of Maxilla (% of LM)	88.15	90.70	10.79	7.55	0.45 / 0.39 / 0.52
WS	Max Distance from Small Segment to LM	21.99	22.71	2.68	2.06	0.27 / 0.34 / 0.36
WS%	Position of  Max Distance from Small Segment  to LM (% of LM)	89.05	91.40	11.25	8.24	0.50 / 0.46 / 0.61
WL	Max Distance from Large Segment to LM	21.91	22.96	2.35	1.75	0.10 / 0.12 / 0.11
WL%	Position of  Max Distance from Large Segment  to LM (% of LM)	86.85	88.90	10.97	9.24	0.57 / 0.53 / 0.50
LLS	Total Alveolar Arch Length	105.27	100.99	8.01	6.83	0. 08 / 0.08 / 0.04
AF	Angle between the vector from Frenulum to the Midpoint between the Sillmann points, and LM	-0.07	1.14	3.18	2.33	0.22 / 0.18 / 0.33

ASIS	Angle between the vector SIS-ALVS and the vector SIS-SID	50.75	49.21	4.50	3.36	0.28 / 0.23 / 0.33
ASID	Angle between the vector SID-ALVD and the vector SID-SIS	70.71	67.21	4.98	4.99	0.05 / 0.03 / 0.05
RWL	Ratio between the posterior width (SID-SIS) and the maxillary length LM	1.04	1.15	0.17	0.16	0.07 / 0.04 / 0.05
RAP	Ratio between the anterior width (at the level of ALVS) and the posterior width (at the level of SID-SIS)	0.50	0.45	0.10	0.08	0.08 / 0.07 / 0.19
RAPs	As RAP, but distances neasured only from small segment to maxillary midline	0.42	0.37	0.13	0.11	0.23 / 0.26 / 0.27

Table 1. Statistical results. Numbers in mm.  p_paired(t) / p(t) / p(r) / p(F) denote p-values of paired t-test / un-paired t-test /
Wilcoxon test / F-test, respectively.
 
 
 

Figure 4.  Plots of distances in mm between the spline curves representing mean maxillary arch forms in Institution A
and Institution B plaster casts, respectively.  The four plots show distances in the x-, y-,  z-directions as well as a
resultant 3D-distance, as indicated.  The x-axis in the plots represent position, in percentage of total arch length,
measured along the arch starting at the SID point.  The vertical dashed line in the plots indicates the position of the
division between large and small segments.
 
 
 

CONCLUSION

In the present paper, the ability of the 3D CT scan analysis method to extract mean arch forms has been demonstrated.
Some significant differences in arch dimensions between Institution A and B, probably caused by the different
treatment protocols, were found. Further analysis is needed in order to investigate the differences in arch form, their
significance, and clinical implications.
 
 

REFERENCES

Bacher M, Goz G, Pham T, Bacher U, Werner O, Buchner P. Three-dimensional analysis of cleft palate topology in
newborn infants with reference to the craniofacial skeleton. Cleft Palate Craniofac J. 1998;35:379-95

Ball JV, DiBiase DD, Sommerlad BC. Transverse maxillary arch changes with the use of preoperative orthopedics in
unilateral cleft palate infants. Cleft Palate Craniofac J. 1995;32:483-8

Berkowitz S. Cleft Lip and Palate. With an Introduction to Other Craniofacial Anomalies. Perspectives in
Management. Singular Publishing Group, Inc. San Diego. 1996;Vol.I:1-388

Berkowitz S. The complete unilateral cleft lip and palate: serial three-dimensional studies of excellent palate growth.
In: Bardach J, Morris HL (eds.), Multidisciplinary Management of Cleft Lip and Palate. Philadelphia: WB Saunders,
1990:456-73

Cyberware product information on the Wold Wide Web 1999. http://www.cyberware.com/products/3030.html

Ferrario VF, Sforza C, Schmitz JH, Colombo A. Quantitative description of the morphology of the human palate by a
mathematical equation. Cleft Palate Craniofac J. 1998; 35:396-401

Foong KWC, Sandham A, Ong SH, Wong CW,  Wang Y, Kassim A. Surface laser scanning of the cleft palate
deformity - validation of the method. Ann Acad Med Singapore. 1999;28:642-9

Friede H, Katsaros C. Current knowledge in cleft lip and palate treatment from an orthodontist's point of view. J Orofac
Orthop. 1998;59:313-30

Gorlin RJ, Cohen MM Jr., Levin LS. Syndromes of the Head and Neck, Oxford University Press. 1990:693-700

Hermann, NV, Darvann, TA, Huebener, DV, Nissen, RJ, Kane, AA, Schlesinger, JK, Kreiborg, S, Marsh, JL(d).  A
method for 3D  shape description of the maxillary alveolar arch.  Poster presented at the 56th Annual Meeting of the
American Cleft Palate - Craniofacial  Association 1999 (abstract)

IDL, Research Systems Inc., Boulder, Colorado, USA, Version 5.0. 1997

Kilpelainen PVJ, Laine-Alava MT. Palatal asymmetry in cleft palate subjects. Cleft Palate Craniofac J. 1996;33:483-8

Kriens O. Data-objective diagnosis of infant cleft lip, alveolus, and palate. Morphological data guiding understanding
and treatment concepts. Cleft Palate  Craniofac J. 1991;28:157-68

Kuijpers-Jagtman AM. Maxillary arch dimensions from birth to 18 months  in complete unilateral clefts. J Dent Res.
1983;62:461-7

Kuijpers-Jagtman AM. Model analysis comparison of BCLP data and the Leighton-data measured with the
OPTOCOM.  In: Kriens O (ed.). What is a Cleft Llip and Palate? A multidisciplinary Update. Georg Thieme Verlag,
Stuttgart.1989:137-9

Lee ST. New treatment and research strategies for the improvement of care of cleft lip and palate patients in the new
millennium. Ann Acad Med Singapore. 1999;28:760-7

McNeil CK. Congenital oral deformities. Br Dent J. 1956;101:191-8

Research Systems Inc., Boulder, Colorado, USA. Using IDL. Version 5.0. 1997

Stellzig A, Basdra EK, Hauser C, Hassfeld S, Komposch G. Factors influencing changes in maxillary arch dimensions
in unilateral cleft lip and palate patients until six months of age. Cleft Palate Craniofac J. 1999;36:304-9

Van der Linden FPGM, Boersman H, Zelders T, Peters  KA, Raaben JH. Three-dimensional analysis of dental casts by
means of the OPTOCOM. J Dent Res. 1972;51:1100-7

Winters JC , Hurwitz DJ. Presurgical orthopedics in the surgical management of unilateral cleft lip and palate. Plast
Reconstr Surg. 1995;95;755-64