A Study of the Use of Photographs for Denture Tooth Selection
Banoo Bindra, BDS, MSc, FDS RCS Ed, Specialist Registrar in Restorative Dentistry, Leeds Dental Institute, United Kingdom, Robin M. Basker, DDS, MGDSRCS, FDSRCS Ed, Professor of Prosthetic Dentistry, Leeds Dental Institute, United Kingdom, John N. Besford, BDS, PhD, MFGDP (UK), Specialist Practitioner, London, United Kingdom.
Reprint requests: Dr B. Bindra, Department of Restorative Dentistry, Leeds Dental Institute, Clarendon Way, Leeds LS2 9LU, United Kingdom. Fax: + 44 113 2336129. e-mail: firstname.lastname@example.org
This study was presented as a poster presentation at the British Society for the Study of Prosthetic Dentistry's Annual Scientific Meeting, March 1999, Liverpool, United Kingdom.
Purpose: The aim was to examine a method described by Wehner et al for calculating the width of a missing central incisor using preextraction photographs. Materials and Methods: Three photographic views were obtained for each of 30 dentate subjects: full face, oblique, and reduced-size full face. The width of the maxillary right central incisor (MR1) was calculated using a formula. The difference between the actual width and calculated width of MR1 was determined for each subject. The median difference and interquartile range were determined because the data were skewed. Results: The width of MR1 calculated using the larger full-face view was typically smaller than the actual width, with a median difference of –0.18 mm. The interquartile range of the difference was from –0.42 to 0.05 mm. For both the oblique and reduced-size views, the calculated width was typically larger, with a median difference of 1.19 mm with an interquartile range from 0.82 to 1.76 mm and a median difference of 0.84 mm with an interquartile range from 0.59 to 1.41 mm, respectively. Conclusion: The technique described by Wehner et al is of proven value in calculating the width of a central incisor when the only available evidence is a preextraction photograph. However, it is of value only when the photograph is a full-face portrait of sufficient size. Int J Prosthodont 2001;14:173–177.
Providing artificial maxillary anterior teeth that closely resemble the patient's missing natural teeth can pose a significant challenge to the dentist, particularly when treatment is sought by patients with extensive tooth loss such that guidance cannot be obtained from the remaining natural teeth. Increasing esthetic awareness and expectations of patients may make this challenge more frequent.
Pound1 stated that five factors--tooth size, form, color,
arrangement, and the framing of the teeth--must harmonize to achieve optimal
facial appearance. Wehner et al2 describe the work of House
and Loop,3 wherein anthropometric data were used in the selection
of artificial teeth. In their study of 555 dentate subjects, it was found
that the greatest bizygomatic width of the skull divided by 16 gave an
approximation of the width of the maxillary central incisor, and when
divided by 3.3 gave an estimated width of the six maxillary anterior teeth
when arranged flat on a card. They also described a method of calculating
the width of a central incisor by measurements made on a preextraction
photograph and on the patient and substituting in the following equation:
where MR1 is the maxillary right central incisor and PD is the interpupillary distance.
Besford4 also drew attention to the value of photographs and noted that the production of good-quality prints had become increasingly common. The various methods that have been described to aid artificial tooth selection have been reviewed by Sellen et al.5 Few of them appear to have been validated by scientific research.
This study aimed to (1) examine the reliability and usefulness of the method described by Wehner et al2; (2) study the effect of photograph size and view on the accuracy of the method; and (3) find out whether using magnifying loupes makes the measurements of the photographs more accurate.
Materials and Methods
The authors obtained ethical approval from the local ethics committee for this study involving a statistically determined sample size of 30 adult dentate subjects. Calculation of the study sample size was based on detection of differences of 0.25 mm at a 5% significance level and 95% power.
Two color photographs of 5 X 7 inches each were obtained for each subject. These showed the full face only, in an anterior view (Fig 1) and in an oblique view approximately 30 degrees to the left (Fig 2). A third color photograph of 5 X 3 inches was obtained, showing the head, neck, and upper body (Fig 3). The width of the head in the third photograph was approximately one fifth that in the larger full-face view. All were taken by the same photographer using predetermined camera settings. Subjects in whom the entire width of the right central incisor was not visible, for example owing to overlapped teeth, were excluded from the study.
The width of MR1 and the interiris distance on the photographs were measured with a ruler to the nearest 0.25 mm. The interiris distance was measured from the most medial point of one iris to the most medial point of the other. This distance, rather than the interpupillary distance, was measured, as it was more accurately identified on the photographs.
For the clinical measurement of the interiris distance, the subject and investigator faced each other. The medial margin of each iris was marked on a wooden tongue spatula, and the distance was measured using the ruler. When making the mark for the subject's right eye, the subject was asked to look at the investigator's left eye and vice versa. This eliminates the convergence of the pupils when focusing on a close object.
All measurements on photographs were made using magnifying loupes with 2.53 magnification. Those of the subjects' teeth and of the distance between the marks on the spatula were made with the naked eye. For a subgroup of 12 subjects, the measurements on the larger full-face photographs were repeated without use of magnifying loupes, and the data were compared.
Three readings were obtained for each measurement, and a mean was calculated.
Repeat readings on the same subject/photograph were made at different
times to avoid memory bias. To calculate the width of MR1, the measurements
were substituted in the following equation:
where ID is the interiris distance. After this calculation was completed, the maximum width of MR1 for each subject was measured directly with a digital caliper (Kennedy Tools). For each of the three photographic views, the calculated width of MR1 was compared with the actual clinical width of MR1. The difference between the two values was determined for each of the 30 subjects. The median difference and interquartile range were determined because the data were skewed.
All measurements were made by the same investigator. Intraoperator variability was assessed for both clinical and photographic measurements. Three measurements of the width of MR1 and the interiris distance of one subject and their photograph were obtained on ten different occasions, totaling 30 readings for each parameter. The mean and standard deviation were determined.
The intraoperator variability was assessed and judged to be within acceptable limits. The standard deviation was found to be well within clinically relevant values (Table 1).
The results for the three photographic views are shown in Fig 4. The width of MR1 calculated using the larger full-face view was typically smaller than the actual width, with a median difference of –0.18 mm. The interquartile range (where 50% of the readings lie) of the difference was from –0.42 to 0.05 mm. Ninety percent of the readings lay within a range from 0.95 mm smaller than the actual width of MR1 to 0.18 mm larger. Using this view, there was a greater tendency to underestimate the width of the tooth than to overestimate it.
For the oblique view, the calculated width was typically greater than the actual width of MR1, with a median difference of 1.19 mm; the interquartile range of the differences was from 0.82 to 1.76 mm.
The median difference between the calculated and actual width of MR1 for the reduced-size photograph was 0.84 mm, the calculated width being typically larger. The interquartile range of the differences ranged from 0.59 to 1.41 mm.
For both the oblique and reduced-size views, there was a marked tendency to overestimate the width of the tooth. The errors were also larger than for the full-face view. The percentage of readings for each view that fell within a range of ± 0.25 mm, ± 0.5 mm, and ± 1 mm when compared with the actual clinically measured tooth width is shown in Table 2. Comparison of the data obtained with and without the use of magnifying loupes did not reveal a difference between the two techniques (Table 3).
The most appropriate length of an artificial tooth is determined by various factors, mainly esthetic considerations related to the desired amount of tooth display. Tooth wear and gingival recession may change the length of the clinical crown, while lip length and skeletal pattern will affect the length of tooth displayed during functional activities. Both lengths are therefore variable. However, the width of an anterior tooth is at least as important as the length, as it helps to establish the visual mass of the tooth and, barring accidents, is more constant than the length.
The results of this study indicate that preextraction photographs are capable of providing a guide to the selection of denture tooth width; however, their usefulness is dependent upon the size and view of the photograph available. For the method to be of acceptable accuracy, the photograph must be in focus, with the maxillary central incisor visible and the eyes open sufficiently to show their structure. Often, only small sizes or views other than the frontal full-face view are available; hence, the method was tested on two different photograph sizes and views.
The interiris distance was used, as it is a relatively constant dimension, with little change after the cranial base has reached its full size in adulthood. Some means of measuring the patient's interiris distance is required; a wooden tongue spatula was used in this study to mark the points, and the distance between them was then measured with a ruler. Special devices used by opticians are also available but are expensive. Pointed dividers, although readily available in most dental surgeries, are potentially dangerous and should be avoided. Measuring one central incisor as opposed to both central incisors (as suggested by Besford4) probably increases the error slightly, but it allows the method to be used when the teeth are spaced or overlapped. The results did not indicate an advantage in the use of magnifying loupes; however, owing to the small sample size for this subgroup, it is likely that very small differences would not be revealed.
The clinical relevance of the technique investigated in this study can be inferred from Table 2. For example, if the clinician would be satisfied with a calculated result that would provide a central incisor that was ± 0.5 mm of the patient's natural tooth, the full-face view of the size used here would be acceptable in 90% of cases. If greater accuracy were sought (± 0.25 mm), it would be possible in 63% of cases. The oblique view is clearly of very little clinical value, and the reduced-size view is little better.
Thus, in a clinical situation, a patient may present with missing maxillary anterior teeth or be edentulous, but they may be able to provide a preextraction photograph. This could then be used to select a maxillary central incisor tooth that resembles closely the missing natural tooth both in size and form. The ratio of the true and photographic interiris distance allows one to scale the photograph and enables calculation of the true central incisor width. Most mold charts available for selecting artificial teeth print the width of the central incisor for a group of anterior teeth. The dentist can then select the group closest to the calculated width of the central incisor and also use the photograph to provide an indication of tooth shape.
The authors are grateful for the expertise provided by the Medical and Dental Illustration Unit of the Leeds Dental Institute, for the statistical assistance provided by Mr Brett Scaife, University of Leeds, and for Mr S. Fayle's assistance with preparation of Fig 4.