Evaluation of radiographic magnification in lateral cephalograms obtained with different X-ray devices : Experimental study in human dry skull

Lateral teleradiographs present extreme relevance in Orthodontics as they demonstrate the relationship of the bone bases and dental relations, besides providing indispensable measures to the orthodontist in growth analysis, diagnosis, treatment planning, therapy monitoring and evaluation of results. Some factors that have great influence on image formation in teleradiographs are: Radiographic distortion, radiographic magnification and errors in the delimitation of the measurement points and cephalometric variables. The objective of this study was to evaluate the influence of radiographic image magnification factor in angular, linear and proportional measurements in cephalometric tracing. Fourteen lateral cephalograms were obtained from a dry human skull with metallic spheres of predefined dimensions, in different devices from three manufacturers: Panoura, Instrumentarium and Tomeceph. After obtaining the teleradiographs, they were scanned. The cephalometric points were marked on the radiopaque images regarding metallic spheres, using the program Radiocef. The program obtained the angular and linear cephalometric measurements. Pearson’ correlation test was used to verify the relationship between the rate of radiographic magnification and cephalometric measurements analyzed. The results confirmed the small change that the angular and proportional measures are subjected to when varying the magnification factor. Linear measurements derived from cephalometric points located in the median sagittal plane presented dimensional changes similar to linear measurements which have as reference the points located in the lateral region of the face. It was observed a high positive correlation between the magnification of the images, regardless of the location of the cephalometric points that generated these measures. The teleradiographs performed on different devices of the same brand did not represent uniformity of measurements. The authors emphasized that it should be avoided the comparison of linear cephalometric measurements from different x-ray machines, even from the same brand and model. Or, alternatively, it should be used a ruler allowing to standardize the magnification rate between radiographs taken on different devices. online article*


17.e2
Evaluation of radiographic magnification in lateral cephalograms obtained with different X-ray devices: Experimental study in human dry skull original article Figure 1 -Diagram showing the process of radiographic magnification.In the position "A" of the film due to the proximity of this one in relation to the patient, there will be less image magnification comparing to the position "B".(Source: Weems, 27 1995).

introduction
Lateral cephalometric radiographs are important in growth analysis, diagnosis, treatment planning and results.The basic equipment to obtain them consists of an X-ray source, an adjustable cephalostat, a chassis for the X-ray film with intensifying screens and the support to the chassis.All these components are rigidly attached to each other and arranged at predetermined distances, thus creating a unit.
For obtaining reliable images, one must know, determine, and if possible, control the several variables related to its performing.Freitas et al 11 pointed didactically the many factors related to the formation of radiographic image.One relates to the geometry and it is also related to the principles of formation of radiographic images, being directly related to the position of the emitting source of X-rays, the object and the film, that are governed by the principle of geometric optics.
[8]11,12,[16][17][18][19]23 radiographic distortion In obtaining lateral cephalometric radiographs, the cephalostat allows the midsagittal plane of the individual to remain parallel to the radiographic film and perpendicular to the X-ray beams.Horizontal head rotations cause changes in the alignment between beams and the object, causing distortions in the image, consisting of the mistaken duplication of a particular structure or area, causing inaccuracies in the analysis of overlaps that could be performed. 22

radiographic magnification
Because the X-rays emanating from the source have a divergent pattern, there is a variation in the amount of magnification of the object in any radiograph. 2,22To reduce the magnification in lateral cephalometric radiographs, one should increase the distance between the source of X-rays and the object to be radiographed in order to take advantage of the central beam, which is flatter, and also decrease the distance between the object and the radiographic film. 4,9,15,21It is recommended a distance of 152.4 cm between the X-ray source and the sagittal plane, considering that increasing the distance would result in loss of penetration of rays. 6According to Weens 27 magnification of craniofacial structures varies from almost 0% up to 24% in objects close to the film or objects in the exact center of the rays.This magnification is not constant for all possible sagittal plane of patient.Structures located closer to the film will present lower magnification comparing to those closer to the rays.As mentioned earlier, another variable, considering the magnification factor, would be the distance between the midsagittal plane of the individual and the film.To minimize variations between different patients and obtain consistent measurements in an individual over time, it is recommended to maintain constant this distance.An average distance of 15 cm is often used, although it would be ideal to position the frame as close to the patient's head as possible to reduce the magnification (Fig 1).

Errors in tracing landmarks and cephalometric measurements
In a study by Bjork and Solow, 7 it was observed that the tracing process introduces significant systematic errors.Depending on the anatomical structures assessed, there is more or less variability in the tracing of landmarks. 5he reproducibility in the tracing of landmarks and linear measurements was studied by Midtgard et al. 17 According to the authors, measurement errors are directly dependent on the precise tracing of the landmarks.In a similar study, Savage et al 21 found out that the reproducibility is dependent upon and varies according to the

17.e3
original article Rino Neto J, Paiva JB, Queiroz GV, Attizzani MF, Miasiro Junior H evaluated structure, and it was not associated to the level of experience of the observer or the quality of the radiographic image for tracing.
The cephalometric radiographs are used by both clinical orthodontists and researchers, that through angular, linear and proportional measurements seek to describe the craniofacial morphology and compare the results of different orthodontic treatments.In order to increase the accuracy in the interpretation of cephalometric measurements, the objectives of this study are: a) To verify the possible influence of radiographic magnification on angular, linear and proportional measurements, formed by anatomical points located in the midsagittal plane or by a combination of points in the midsagittal plane and lateral regions of the face.
b) To check the rates of radiographic magnification in lateral cephalometric radiographs obtained in different machines from the same manufacturer.

MAtEriAL And MEtHodS
The sample consisted of 14 lateral cephalometric radiographs of a dry skull belonging to the Institute of Biomedical Sciences, University of São Paulo, obtained using 14 different X-rays devices.The device models provided by radiological institutes were Panoura (10CSU-Yoshida, Japan), Instrumentarium (OP100, Instrumentarium, Finland) and Satelec/Tomeceph (XMindSatelec / Tomeceph-Orion Corp., Finland).
The position of the skull was oriented in the cephalostat, with the Frankfort horizontal plane parallel to the horizon, while maintaining the midsagittal plane perpendicular to it.The ear rods were inserted into the external auditory canal, preventing rotation of the skull and then adjusted to nasal positioner for maintenance of the position (Fig 2).To standardize the acquisition of the radiographs, only devices whose chassis were positioned on the right side of the skull were used.The upper teeth were fixed to the lower ones with acrylic resin to maintain constant the relation between the mandible and the skull, avoiding errors in the measurements.Due to the difference between the specifications of the devices, their adjustment was based on the quality of the radiographic image produced, especially in the sharpness of radiopaque markers.The brand of films used were Kodak, 18x24 T-MAT G/RA.The radiographic processing was performed in the radiology institutes themselves, always using automatic processors.
In order to eliminate errors in the identification of landmarks that could lead to errors of reproducibility and measurement, 5,7,12,14,16,17,19 radiopaque markers were fixed in dry skull (standardized steel balls with 1.0-mm diameter) at the sites described in Table 1 and shown in Figures 3, 4 and 5.In relation to points located outside the midsagittal plane, it was advocated fixing the metallic markers on the right side of the skull to make them closer to the chassis and to have the smallest magnification. 22fter obtaining the lateral cephalometric radiographs, they were scanned.The landmarks were traced on the metal spheres, using the Radiocef program (Ra-dioMemory-Brazil).The lines and plans shown on Figure 6 and Table 2 were outlined for evaluation.

Nasion (N)
Anterior-most point of frontonasal suture Anterior sella (AS) The deepest portion of the anterior concavity of the sella turcica.
Anterior nasal spine (ANS) Vertex of the anterior nasal spine.
Posterior nasal spine (PNS) Vertex of the posterior nasal spine.

Point A
The deepest point of the anterior concavity in the maxilla.

Menton (Me)
Most inferior point of the mandibular symphysis.

Gonion (Go)
Most posterior inferior point on the mandibular angle.

Xi
The geometric center of the ramus of the mandible

17.e4
Evaluation of radiographic magnification in lateral cephalograms obtained with different X-ray devices: Experimental study in human dry skull original article using the Dahlberg's formula.To evaluate the error of measurement as to the tracing of the landmarks, five lateral radiographs were randomly selected and submitted again to the digitization process.
The statistical descriptive analysis was performed and the Pearson correlation test was used to verify the relationship between the rate of radiographic magnification and the accuracy of cephalometric measurements.Table 5 -Proportion measurements.
ANS-PNS:Sa-N Proportion between the distances ANS-PNS and As-N.
Formed by lines located in the midsagittal plane.

Sa-N:Go-Me
Proportion between the distances Sa-N and Go-Me.
Formed by a line located in the midsagittal plane and one on the right side of the midsagittal plane The measurements were obtained by a single operator, twice, at different times, and their means were considered for analysis.The angular and linear cephalometric measurements were presented as means, medians and standard deviations.To evaluate the accuracy in the positioning of the skull in the cephalostat, two new lateral cephalometric radiographs were obtained in five of the devices used.These measurements were compared original article Rino Neto J, Paiva JB, Queiroz GV, Attizzani MF, Miasiro Junior H measurements showed high positive correlation for linear measurements, while the angular and proportional did not correlate (Table 7).The radiographs obtained with each brand of equipment were ordered in increasing order of magnification from the lowest value found to Sa-N linear measurement.The highest differences were of 0.6 mm, 0.8 mm and 1.5 mm, respectively for the brands Instrumentarium, Panoura and Satelec.
In Table 8 one can observe the percentage values of corresponding magnifications.

diScuSSion
The reliability of the results of this study depends on the control of errors related to the positioning of the skull in the tracing of landmarks.

rESuLtS
There were no significant differences in the evaluation of the method error according to data obtained with the Dahlberg's formula.Considering the positioning of the skull and the tracing of landmarks, the method error was 0.27 mm for linear measurements, 0.27° for angular measurements, and 0.0% for proportion measurements.The general values obtained for cephalometric measurements are shown in Table 6.The radiographs were ordered in increasing order of magnification from the smallest value found to Sa-N linear measurement.
Pearson's linear correlation test to investigate the correlation between the rate of magnification of radiographic images and the accuracy of cephalometric

17.e6
Evaluation of radiographic magnification in lateral cephalograms obtained with different X-ray devices: Experimental study in human dry skull original article As found in the literature, a frequent cause for the appearance of ghost images is related to improper positioning of the patient in the cephalostat. 3,4,18,22,23Studies about radiographic distortion using dry skulls are mentioned in the literature, evaluating the influence of head position in radiographs acquisition.According to Ahlqvist et al, 3 rotations of the object up to 5° resulted in a variation of less than 1% in length of the linear measurements.Rotations greater than 5° may result in noticeable cephalometric errors, and as they are evident when positioning the patient, they must be corrected prior to exposure.In a study by Spolyar, 22 it was observed a mean linear change of 1.7 mm ranging from 0.5 to 6.2 mm, and mean angular changes of 1.59°, ranging from 0 to 5.23°.According to another study by Ahlqvist et al, 4 cranial rotations of 5° to 10° were responsible for significant distortions in the angular measurements.Rotation of the head in the vertical direction (Z axis) also cause a distortion, and Yoon et al 28 noted that the angular measures showed projection errors smaller than the linear measures, and that the use of references located on the midsagittal plane for angular measurements showed less distortion.
In the present study to evaluate the existence of errors related to the positioning of the skull in cephalostat, two lateral cephalometric radiographs were taken at different times, totaling five X-ray devices in the evaluation of the method error.There were no rotations of the skull, discarding the incorporation of significant distortions that could influence the cephalometric measurements.
Another problem for studies that use cephalometric radiographs, is the location of the landmarks. 5,7,13,17,20,21,23ts accuracy depends on the anatomical structure used, the radiographic image quality, visual acuity and experience of the operator on locating the landmarks, considering that great care must be taken so that measurement errors do not influence the study. 5,7In order to avoid the incorporation of errors in the identification of landmarks, we chose to fixate 1-mm radiopaque spheres in selected regions of the dry skull.
The first objective of this study was to evaluate the influence of image magnification rate on the angular, linear and proportional cephalometric measurements using lateral cephalometric radiographs.Measures located in the midsagittal plane and lateral regions of the face were analyzed.
According to Bergensen,6 the magnification rate in lateral cephalometric radiographs ranges from 4.6% to 7.2%.The results are related to the location of landmarks outside the midsagittal plane and to the variation of distance between the film and the X-ray source.Adams 1 in its cephalometric study also noted a significant variation for linear measurements, as the distance from the X-ray source was increased.In the same study, the angular measurements showed an average increase of 1°, with less variation for measures which references were in the midsagittal plane, and greater variation for those located in the mandible.The author concluded that the variation for these measures, especially the linear ones, increases as the distance from the central beam of X-rays is increased.
According to the results obtained with the Pearson's test, the linear measurements showed a high positive correlation with the rate of image magnification, while the angular and proportional ones, did not correlate (Table 7).These results confirm that the angular measurements are subjected to small changes.This was also observed for proportional measurements, which showed small correlation.
Linear measurements derived from landmarks located in the midsagittal plane showed dimensional changes that were similar to those from linear measurements that have reference points located in the midsagittal plane and in the lateral regions of the face (Table 6).
The magnification rate of radiographic images is an important factor because most orthodontic studies are based on cephalometric analysis and its measurements are made on cephalometric radiographs obtained with different X-ray devices. 12,14However, despite it is known that it influences the measures, the magnification rate is not mentioned in several articles.Radiographs collected for several years are an important source for retrospective studies; however, not knowing its magnification rates prevents its use in longitudinal studies.A possible alternative to validate the use of these radiographs would be to ensure equal magnification rates in devices with the same brand and model, allowing for the determination of the magnification rate in equal devices still in use.
The second objective of this study was to determine the amplitudes of radiographic magnification on lateral cephalometric radiographs obtained with different X-ray devices from the same brand and model.

17.e7
original article Rino Neto J, Paiva JB, Queiroz GV, Attizzani MF, Miasiro Junior H The images obtained with different X-ray devices of the same type and brand did not show uniform magnification rates, with linear variations ranging from 0.4 to 1.5 mm, corresponding to magnification percentages of 0.6 to 2.3% (Table 8).
Although variations in the magnification rate of the device of the brand Instrumentarium were very low, generally it is not recommended to compare linear cephalometric measurements made with different X-ray devices, even those of the same brand and model, being necessary to use a ruler to standardize the magnification rate of radiographs.

concLuSionS
The radiographic magnification showed high positive correlation with the linear measurements, being responsible for significant variations, regardless of whether the anatomical points were located in the midsagittal plane or lateral regions of the face.The angular and proportional measurements showed no significant correlation.
Performing cephalometric radiographs with different devices of the same brand and model does not ensure uniformity in the rates of radiographic magnification.

Figure 2 -
Figure 2 -Positioning the skull on the cephalostat.

Figure 3 -
Figure 3 -View of the radiopaque marker located in the mandibular angle.

Figure 4 -
Figure 4 -View of the radiopaque marker located in the frontonasal suture, corresponding to the point N.

Figure 5 -
Figure 5 -Scheme of the dry skull, with the representation of the sites where the radiopaque markers were attached.

Figure 6 -
Figure 6 -Scheme of the dry skull with the cephalograms for representation of reference lines and planes.

Table 2 -
Lines and planes.

Table 6 -
Results of linear, angular and proportional measurements obtained with different devices.

Table 7 -
Pearson correlation test to evaluate the correlation between the magnification and accuracy of measurements.

Table 8 -
Results of linear measurements obtained with different devices of the same brand and model.