Stereo-assisted landmark detection for the analysis of changes in 3-D facial shape

In this paper, a semi-automated approach to 3-D landmark digitization of the face is described which uses a combination of active shape model-driven feature detection and stereophotogrammetric analysis. The study aims to assess whether the proposed method is capable of detecting statistically significant changes in facial soft tissue shape due to mandibular repositioning in a cross-sectional patient sample. A hybrid stereophotogrammetric and structured-light imaging system is used for acquiring 3-D face models in the first instance. A landmark-based statistical analysis of facial shape change is then carried out using procrustes registration, principal component analysis and thin plate spline warping on the 2-D facial midline profiles and automatically digitized 3-D landmarks. The proposed method is validated both statistically and visually by characterizing shape changes induced by mandibular repositioning in a heterogeneous cross-sample of 20 orthodontic patients. It is shown that the method is capable of distinguishing between changes in facial morphology due to simulated surgical correction and changes due to other factors such as growth and normal variation within the patient sample. The study shows that the proposed method may be useful for auditing outcomes of clinical treatment or surgical intervention which result in changes to facial soft tissue morphology.

Proportional changes in cephalometric distances during Twin Block appliance therapy

The aim of the study was to evaluate the cephalometric changes produced by the Twin Block appliance. Lateral cephalometric radiographs were taken before and after Twin Block appliance treatment on 30 consecutive patients (14 male, 16 female, mean age 12 years 6 months). Published normative data tables were used to produce control data, which were individually matched to the test group for age, sex, and treatment time. Alteration in shape was assessed by measuring percentage change in linear dimensions as opposed to change in cephalometric angles used in previous investigations. The differential average percentage change was calculated by subtracting the control value from the Twin Block value. Clinically significant changes (2 per cent and greater) were found in lower anterior (6.6 per cent) and posterior (4.6 per cent) face heights, upper incisor to maxillary plane (4.9 per cent), i.e. upper incisor retraction, and increase in mandibular length (Co-B 3.3 per cent, Co-Po 2.6 per cent, Ar-B 3.5 per cent, Ar-Po 2.2 per cent).

A cephalometric evaluation of the Biobloc technique using matched normative data

PATIENTS AND METHOD

The outcome of Biobloc treatment was evaluated on 35 patients, 13 male, 22 female, with Class II, Division 1 malocclusions. Cephalometric radiographs were taken before and after treatment. A control group was generated from published normative data in such a way that each treated case was matched individually for age, sex and treatment time. The cephalometric change during treatment was compared to the natural growth change in the matched control group using a Mann-Whitney U-test. The treatment effect was calculated by subtracting the natural growth change from the treatment change. This was then compared to twice the method error to see if the treatment change was clinically significant.

RESULTS

There was both a statistically and a clinically significant reduction in overjet (p < 0.001), overbite (p < 0.001) and angle ANB (p < 0.001) with an increase in angle SNB (p < 0.001). None of the other cephalometric parameters measured showed a significant change except for upper incisor proclination (p < 0.05) and ArGoMe angle (p < 0.05) but these were not clinically significant.

Cephalometric evaluation of the Twin-block appliance in the treatment of Class II Division 1 malocclusion with matched normative growth data

Cephalometric radiographs were taken before and after treatment with the Twin-block appliance on 30 consecutive patients with Class II Division 1 malocclusions. A control group was generated from published normative data such that each treated case was matched individually for age, sex, and treatment time. The cephalometric change during treatment was compared to the natural growth change in the matched control group using a Mann-Whitney U-test for statistical significance. The treatment effect was also calculated by subtracting the natural growth change from the treatment change. This was then compared to twice the method error to see if the treatment change was clinically significant. There was both a statistically and clinically significant reduction in overjet, angle ANB, increase in angle SNB, and mandibular length together with a reduction in upper incisor angle. None of the other cephalometric parameters measured showed a significant change.

Relationship of the functional oropharynx to craniofacial morphology

The association between the functional oropharyngeal airway (defined as the minimal sagittal dimension at right angles to the airstream) and craniofacial morphology was investigated using 16 craniofacial variables taken from lateral cephalometric radiographs. The sample consisted of 70 subjects (31 males and 39 females) 10 to 13 years of age. There was no difference in ages between males and females, and no correlation with age except upper face height. Oropharyngeal airway was positively correlated with length of the mandible (Gon-Men), the distance between the third cervical vertebra and the hyoid bone (C3-Hy), and cranial base angle (NSBa). Although short mandibular length is a characteristic finding in patients with obstructive sleep apnea, none of the subjects in this study had this diagnosis.

Differences in cephalometric reference values between five centres: relevance to the Eurocleft Study

Cephalometric data from five different geographical areas (Oslo, Manchester, London, Nijmegen, and Michigan) were compared. The angles SNA, SNB, and ANB were taken from published tables in the form of mean, standard deviation (SD) and number in the sample, divided up by age and sex. Angles SNA and SNB were significantly different among centres on univariate analysis. The Oslo data showed 23 instances in which they differed significantly from the other data, London 16, Manchester 11, Nijmegen 11, and Michigan 7. Multivariate analysis using Penrose distances were Oslo (2.04), Manchester (1.39), London (0.93), Nijmegen (0.80), and Michigan (0.66). The sums of the Mahalanobis distances were Nijmegen (3.60), Oslo (3.10), London (2.80), Manchester (2.25), and Michigan (1.49). As these results probably reflect racial and genetic differences, these must be taken into account when international comparisons are being made, as in the Eurocleft study.

Questionnaire evaluation of feeding methods for cleft lip and palate neonates

A questionnaire evaluation was undertaken of feeding methods used by the mothers of 25 neonates with cleft lip and/or palate. Most parents had problems feeding their babies, both with the quantity of food taken and especially with the time taken to feed; even after a period of 2 months over a quarter were still having problems with the quantity of feed and a third had not established a regular feeding pattern. Just over a half needed to change the feeding method from the one with which they started. Twelve of the 25 mothers tried to breast-feed but none of them was successful. The Haberman feeder was the most popular, being used by 18 mothers. An acrylic feeding plate was considered to be helpful by 6 of 11 mothers of babies with complete clefts but by only 1 of 9 mothers of babies with secondary palate clefts. Almost all the mothers were dissatisfied with the information they had received while in hospital and with the back-up when they went home.

Alteration in facial shape in the anencephalic human foetus

Tracings were produced from frontal radiographs of 11 anencephalic foetuses. These were compared analytically with standards derived from 60 normal foetuses. The anencephalics showed a marked reduction in the relative size of the cranium with considerable flattening of the calvarium, which is characteristic of the condition. The orbits were more medially placed, being closer together and somewhat higher in position that the normal foetal outline. The right and left mandible were more elongated with an increase in the intermaxillary space. The condyles were more medially positioned and were closer to the orbits than in the normal outlines. In normal growth, expansion of the brain flattens the cranial base, displacing the nasomaxillary segment forwards. In anencephally this does not happen so that the nasomaxillary segment rotates in an antero-inferior postero-superior direction along with the intermaxillary space and mandible. This produces a significant lengthening of the face accompanied by narrowing in the region of the cranial base relative to the normal outline.

Relative exposure of children to lead from dust and drinking water

The Blackpool, Wyre and Fylde Health Authority, in the North West of England, could be described as a "low-level lead exposure area." Primary sources of lead exposure are atmospheric fallout (both indoors and outdoors) and potable water consumption. Deciduous teeth were collected from children living in this area as were water samples and outdoor dust samples. Both total lead concentrations and 206Pb:207Pb ratios were determined for a defined subset of teeth. Significant differences in the total lead concentrations were found for teeth collected from children resident in different targeted areas (i.e., Blackpool, Fleetwood, and Garstang). No significant differences were found between the total lead concentrations or the 206Pb:207Pb ratios from dust and water samples in these areas. Examination of the 206Pb:207Pb ratios for dust, water, and teeth obtained from each area separately revealed differing patterns of exposure to lead. Determination of 206Pb:207Pb ratios, in addition to total lead concentrations, enabled the differences in sources of exposure to be identified in these communities. The authors conclude that isotopic analyses are an important aspect of community survey work, and these analyses can be helpful in accurately targeting intervention strategies aimed at reducing exposure to lead.

A comparison of Twin Block, Andresen and removable appliances in the treatment of Class II Division 1 malocclusion

Cephalometric radiographs were taken before and after treatment in three groups of patients. Ten were treated with removable appliance therapy involving extraction of upper first premolars and retraction of anterior teeth. Ten were treated on a non extraction basis coupled with wearing an upper expansion appliance followed by an Andresen appliance. Another group of ten patients were treated in an identical manner except that they wore a Twin Block instead of the Andresen appliance. When comparing the patient groups before and after treatment, statistically significant differences were found for upper incisor angulation in the Removable (t = 9.49 p < 0.001) Andresen (t = 4.48 p < 0.001) Twin Block (t = 2.81 p < 0.001) groups and for angle ANB in the Andresen (t = 2.88 p < 0.01) and Twin Block (t = 2.65 p < 0.01) groups. Thus the overjet was reduced purely by incisor tipping with the removable appliance as would be expected, but both functional appliances produced some bodily correction with an improvement in the dental base relationship. Although the Twin Block appliance group showed a greater reduction in angle ANB this was not significantly different from that in the Andresen appliance group. The treatment time from insertion of the first appliance to reduction of the overjet to 3mm or less was also recorded. The treatment time for the Twin Block group (8.1 months) was significantly less (t = 2.79 p < 0.05) than the Andresen group (12.8 months), but neither functional group differed significantly from the removable appliance group (10.1 months).

Relative growth of the human fetal skull in width, length and height

Linear measurements of the width, length and height of the face and cranium were made on 60 fetuses of 49-212 mm crown-rump length. Absolute growth rate was derived from the slope of the regression line of dimensional length against time. There was a high correlation between absolute growth rate and mean dimensional length (r = 0.995; p less than 0.001). The relative growth rate was calculated by dividing the absolute growth rate by the mean dimensional length. Change in shape was measured by the relative increase or decrease in dimensional length once the outlines were normalized to an equal area to eliminate the effect of size. Change of shape correlated highly with relative growth rate (r = 0.95; p less than 0.001). Relative growth rates provide a reliable indication of change in shape, unlike absolute growth rates, which are mainly influenced by size. Measurement of relative growth rates showed an increase in cranial length greater than height greater than width and facial length greater than width greater than height.

A functional appliance system for the correction of class II relationships

A technique is described for the correction of Class II malocclusion involving a functional appliance and consisting of three phases. 1. Adjustment of posterior arch width and dental alignment, using semi-rapid maxillary expansion by means of an upper removable appliance, to co-ordinate the anticipated positions for the arches. 2. Correction of the Class II dental relationships using a modification of the twin-block functional appliance. 3. Retention, using an upper removable appliance with a very steep anterior bite plane. The technique is illustrated by a case report.

A quantified comparison of craniofacial form with nasal respiratory function

Rhinomanometry provides a means to quantify the nasal airway in terms of its conductive efficiency by use of the inverse or resistance to airflow. In samples of 21 male and 26 female subjects, nasal airway resistance was compared with a number of form measurements and the coefficients of correlation were calculated. Statistical significance was found in some of the relationships with nasal airway resistance, particularly the maxillary-mandibular plane angle, the palate-tongue distance, the palatal width, and the facial index.

Asymmetry of the human skull during fetal growth

Asymmetry of the human skull in the frontal and basal views has been described previously as a normal feature in adults. However, no investigation has been performed during the fetal period to see when asymmetry first develops. Previous quantitative methods for measuring asymmetry have relied upon defining anatomical points and using geometrical constructions. In the present study image outlines and their centroids (centers of area) were related to each other via a rectangular coordinate reference grid using the technique of morphanalysis. The centroids were shifted to the left relative to the coordinate reference grid indicating a left side expansion. Also the signed area under the cyclical curve, which is zero for a perfectly symmetrical shape, showed a negative shift indicating asymmetry. The cranium was found to be the major site of asymmetry caused by unequal development of the cerebral hemispheres. The asymmetrical development of the temporal lobes caused a greater displacement of the nasomaxillary segment on the left leading to a rotation of the face on the cranium.

The relationship between differences in regional growth rates and changes in shape during human fetal craniofacial growth

Large differences in growth rates between various craniofacial regions have been reported but have not distinguished between absolute and relative rates. Measurements on 60 fetuses, 49-212 mm crown-rump length, showed that absolute growth rates correlated highly with size of the craniofacial region measured (r = 0.9996, p less than 0.001) but were unrelated to changes in shape and relative proportion. By standardizing for size, absolute growth rates were converted into relative growth rates which correlated with changes in relative proportion and thus shape (r = 0.785, p less than 0.001).

Changes in the jaw relationships during human foetal cranio-facial growth

Previous investigations into pre-natal cranio-facial growth have either been purely descriptive or have involved the use of cephalometric points, lines and angles. Neither methods have given an entirely clear picture of the changes taking place during the foetal period. The present investigation measured changes in shape, position and orientation of the jaws by using image outlines and their centroids (centres of area) in combination with the analytic method of morphanalysis. The image outlines of the foetuses were related to each other via a rectangular co-ordinate reference grid so that the exact sites where growth changes occurred could be plotted and isolated. It was found that the jaw relationships were constantly changing during the foetal period. Brain growth appeared to predominate which flattened out the cranial base and determined the definitive position of the naso-maxillary segment to which the musculature of the lower face in turn adjusted mandibular growth. The jaw relationships changed in the direction Class III to Class II mainly as a result of forward growth of the naso-maxillary segment, the mandible showing a lag followed by a catch-up regaining its original position. The two jaws exhibited different growth patterns which suggested differences in the underlying mechanisms responsible for their growth. This is almost certainly an important factor in the aetiology of malocclusion resulting from malrelationships of the basal bones and soft tissues.

Shape changes during human fetal craniofacial growth

An investigation into craniofacial growth during the fetal period was undertaken in order to measure changes in shape. Previous workers have reported little or no change in cephalometric angular measurements, leading to the conclusion that facial shape is stable, in contradiction to the changes in shape perceived by simple observation. The major problem has been the development of quantitative methods which accurately measure shape. The analytical method of morphanalysis was used in order to overcome the problem. Image outlines of fetuses were related to each other via a rectangular reference grid. The exact sites where growth changes occurred were plotted in relation to the coordinate reference grid and therefore could be isolated. The shape of the head was found to change progressively in all three dimensions during the fetal period. The impression given by cephalometric studies that facial shape is stable was not supported by the present investigation, which clearly demonstrated a progressive alteration in shape with age. It appeared that brain growth predominated over that of the face during the fetal period. This not only produced an alteration in the shape of the cranium, causing its elongation, but influenced the position of the nasomaxillary segment to which the musculature of the lower face in turn adjusted mandibular growth.

Angular changes in cephalometric and centroid planes during foetal growth

Alteration in shape of the cranio-facial region during the foetal period was measured by recording the angular changes between cephalometric planes and centroid planes. Cephalometric planes rely on locating anatomical landmarks whilst centroid planes join mathematically derived points. Centroid plane angles showed a higher correlation with age than cephalometric plane angles and were demonstrated to be a superior measure of change in shape. Both methods revealed a change in jaw relationship from Class III in the younger foetuses to Class II in the older ones. This was brought about by forward growth of the maxilla in excess of the much smaller forward movement of the mandible.

The relationship between bruxism and temporomandibular joint dysfunction as shown by computer analysis of nocturnal tooth contact patterns

Whilst there are many causes of temporomandibular joint dysfunction, one of the most frequently encountered is bruxism. The two conditions have many aetiological factors in common and there is a rational physiological reason why bruxism could cause temporomandibular joint dysfunction. In fact, experimental bruxism performed under laboratory conditions has been found to cause such joint dysfunction. This contention is supported by direct measurements of nocturnal tooth contact. Patients with temporomandibular joint dysfunction were found to differ from control subjects to a degree which was statistically significant.

The stability of anatomical and centroid reference points in cephalometric analysis

Cephalometric records of 60 fetuses were combined on a coordinate reference grid to measure the statistical spread of the outlines. When centroid instead of anatomical points were used for superimposition of records, the spread of the outlines was reduced dramatically, being statistically significant (p less than 0.001). The centroid is a mean point being less variable and preferable to anatomical points commonly used for comparing successive cephalometric records.