ORIGINAL ARTICLE
Comparison of Orbital Volume Obtained by Tomography and Rapid Prototyping Guilherme Berto Roc¸a, MD,* Jose´ Aguiomar Foggiatto, PhD,Þ Maria Cecilia Closs Ono, MD,þ Sergio Eiji Ono, MD,§ and Renato da Silva Freitas, PhD|| Purpose: This study aims to compare orbital volume obtained by helical tomography and rapid prototyping. Methods: The study sample was composed of 6 helical tomography scans. Eleven healthy orbits were identified to have their volumes measured. The volumetric analysis with the helical tomography utilized the same protocol developed by the Plastic Surgery Unit of the Federal University of Parana´. From the CT images, 11 prototypes were created, and their respective volumes were analyzed in 2 ways: using software by SolidWorks and by direct analysis, when the prototype was filled with saline solution. For statistical analysis, the results of the volumes of the 11 orbits were considered independent. Results: The average orbital volume measurements obtained by the method of Ono et al was 20.51 cm3, the average obtained by the SolidWorks program was 20.64 cm3, and the average measured using the prototype method was 21.81 cm3. The 3 methods demonstrated a strong correlation between the measurements. The right and left orbits of each patient had similar volumes. Conclusions: The tomographic method for the analysis of orbital volume using the Ono protocol yielded consistent values, and by combining this method with rapid prototyping, both reliability validations of results were enhanced. Key Words: Orbit, tomography, diagnostic (J Craniofac Surg 2013;24: 1877Y1881)
A
large number of congenital, vascular, endocrinal, tumoral, and traumatic anomalies are able to change the orbital anatomy. The What Is This Box? A QR Code is a matrix barcode readable by QR scanners, mobile phones with cameras, and smartphones. The QR Code links to the online version of the article.
From the *Federal University of Parana´; †Federal University of Technology of Parana´; ‡Plastic Surgery Unit, Federal University of Parana´; §Department of Radiology of DAPI; and ||Plastic Surgery Unit, Federal University of Parana´, Curitiba, Brazil. Received March 18, 2013. Accepted for publication May 5, 2013. Address correspondence and reprint requests to Dr. Guilherme Berto Roc¸a, Plastic Surgery Unit, Federal University of Parana´, Rua General Carneiro, 181VHospital de Clı´nicas Curitiba, CEP 80060-900, Curitiba, Brazil; E-mail: [email protected] This is a Thesis of Master Degree of G.B.R., Post Graduate of SurgeryVFederal University of Parana´. The authors report no conflicts of interest. Copyright * 2013 by Mutaz B. Habal, MD ISSN: 1049-2275 DOI: 10.1097/SCS.0b013e31829a8332
The Journal of Craniofacial Surgery
variation of orbital volume from several pathologies can lead to profound changes in ocular physiology.1 Many diseases that alter the orbit can be treated by surgical procedures to recover its volume. Surgical planning for these patients encounters difficulties due to lack of accuracy and precision inherent with a majority of the evaluation methods.2 Among acquired diseases affecting orbital volume, Graves ophthalmopathy is one of the most frequent. Its association with hyperthyroidism was first described by Robert Graves in 1835. The disease is well known, with a clinical presentation of proptosis caused by an imbalance in the relationship between the orbital content and continent. Among the traumatic causes, facial trauma with fracture of orbital walls commonly alters the orbital volume, with important clinical implications.3 As the orbit is both a functional and aesthetic unit of the face, abnormality of orbital volume is apparent and significant.4 The concept has already been established that restoration of orbital volume can relieve ocular symptoms.5 The methods for evaluation of orbital volume as currently described are difficult to reproduce, and require specific equipment and training. Moreover, there are marked discrepancies in values assigned by different observers.6 Therefore, in clinical practice, the surgeon must rely on his/ her personal observational and surgical skills. A reliable and accurate method could help in understanding the physiopathology and perioperative management of these patients, and more objective volumetric evaluation could help in long-term postoperative follow-up of previously treated patients.7 Undoubtedly, the development of a new radiological method for measuring orbital volumes with easier implementation, improved accuracy, and low cost would greatly benefit patients with orbital fractures and volumetric changes requiring surgical intervention.
PATIENTS AND METHODS This study compared orbital volumes obtained by helical computed tomography and rapid prototyping, using helical CT scans of 6 patients with no evidence of orbital volume abnormality, from a database of the Radiology Department of DAPI (Advanced Diagnostic Imaging), located in Curitiba, Parana´. Sample selection included examination of the data for patients between 20 and 60 years of age (mean age 36 years) with no anomalies reported and no reported medical conditions that would alter the orbital volume, such as sinus disease and retinal disorders. Patients whose reports showed evidence of ocular ophthalmopathy, malformations, and/or tumors were excluded. Six patients with a total of 11 orbits were eventually selected. One patient showed a change in the left orbital floor, and that orbit was excluded from the project. This study was conducted in 3 centers: the Plastic and Reconstructive Surgery Unit of the Federal University of Parana´, the Department of Radiology of DAPI, where Dr. Sergio Ono was responsible for conducting the volumetric analyses in helical CT scans, and the Federal University of Technology of Parana´, at the Prototyping Nucleus (NUFER), under Prof. Dr. Jose´ A. Foggiatto, responsible for the creation and production of the 11 prototypes.
& Volume 24, Number 6, November 2013
Copyright © 2013 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
1877
The Journal of Craniofacial Surgery
Roc¸a et al
& Volume 24, Number 6, November 2013
FIGURE 1. A, B, The Ono method for analyzing orbital volume using a CT scan.
The orbital volume obtained by spiral CT (Ono method) was performed by a single radiologist (Dr. Sergio Ono), in compliance with previously published protocol.8 Using this method, 17 slices of 0.5 mm of thickness were obtained from the CT scans. Then, with the sum of the consecutive cones, the volume of the orbit was calculated (Figs. 1A, B). To analyze the orbital volume in the prototype, 11 models were created and produced by the Stratasys FDM 2000 machine using copolymer ABS (acrylonitrile, butadiene, and styrene) (Fig. 2). The average time to manufacture each orbit was 3 hours. To measure the orbital volume of the prototypes, they were filled with saline solution. The saline was used as it is a noncorrosive material, easy to handle, and offers volume values in cubic centimeters. A drop of detergent was added to the saline solution to break the surface tension and reduce measurement errors. Otherwise, the volume injected into each orbit could exceed the anatomical boundaries used for landmarking (Fig. 3). During the fabrication process of the prototypes using the InVesalius program, it was possible to make a new assessment for orbital volume analysis using the SolidWorks software (Fig. 4). Therefore, 3 methods were available to measure the orbital volume: the Ono method, SolidWorks analysis, and the prototype method (direct measurement with saline). The 3 methods of orbital volumetric measurements were described by mean, median, minimum values, maximum values, and standard deviations. Data were analyzed with Statistica v.8.0 software.
RESULTS After obtaining the volumes of the 11 orbits, the results were tabulated and compared with volumes obtained by imaging methods (Table 1). We tested the null hypothesis that the mean volume is equal to the 3 orbital measurement methods versus the alternative
FIGURE 2. The prototype.
1878
hypothesis that at least one of the methods has a different average than the others. Table 2 and Figure 5 show descriptive statistics of the volume according to the method and the P value of the statistical test. The results showed no significant difference between the volume measured between the Ono method and the SolidWorks method (P = 0.724). There was, however, a significant difference between the Ono and prototype methods (P = 0.001) and between the SolidWorks and prototype methods (P = 0.003) in comparing average orbital volume. For each method, we tested the null hypothesis that the mean volume on the right side is equal to the average volume on the left side versus the alternative hypothesis of different means. The results in Table 3 and in Figures 6, 7, 8, and 9 demonstrated that with the Ono and the prototype methods, the orbital volumes between left and right orbit are not significantly different; however, with the SolidWorks method, this difference was significant.
DISCUSSION Tomography has been the most frequently used method for measuring orbital volume.1,9 Current protocols for the evaluation of orbital volume are not universally accepted. Every radiology and craniomaxillofacial surgery service uses a method and a standard program for orbital volumetric analysis. However, despite these various available methods of measurement, no single method is ideal. Furthermore, most published studies have not compared their results against other measurement methods. Therefore, it is difficult to know how reliable these results are. Several studies have attempted to compare the volumetric analyses obtained by tomography versus direct volumetric measurement of the orbits of skulls. Cooper et al compared the orbital volume obtained with CT against the volume obtained by filling the orbits of skulls with dry sand.1 The discrepancy between the
FIGURE 3. The prototype filled with saline.
* 2013 Mutaz B. Habal, MD
Copyright © 2013 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
The Journal of Craniofacial Surgery
& Volume 24, Number 6, November 2013
Comparison of Orbital Volume
TABLE 2. Orbital Volume Analysis in the Three Methods Method
n
Average
Minimum
Maximum
Standard Deviation
P*
Ono SolidWorks Prototype
11 11 11
20.51 20.64 21.81
17.66 17.03 18.20
23.40 24.05 24.80
2.29 2.16 2.57
0.002 0.002 0.002
*Analysis of variance with repeated measures, P G 0.05.
FIGURE 4. The SolidWorks method for analyzing orbital volume.
results obtained by the direct method and the tomographic method ranged between 0.2% and 4%. Forbes et al also established comparisons between silicone models obtained using dry skulls and examinations of patients without pathologies affecting their orbits. This work was very helpful in establishing normal values for orbital structures, to be used as standards in future studies. The average for bony orbit volume is about 23 cm3 and the average for intraconal fat is 10 cm3.9 The study of Cooper et al and Forbes et al characterized their tomographic methods as reproducible and with good accuracy; however, in analyzing these studies, there is a simple error: none of them used a gold standard method against which to compare the results, so the claim of objective reliability is somewhat compromised. Various kinds of software have been used to analyze the orbital volume. Regensburg et al used the Mimic (Materialise Inc.) program, which calculates the orbital volume in a manner similar to the method proposed by Ono et al.10 Kwon et al used the Vitrea program, version 3.4 (Vital Images Inc.) and Dextroscope version 1.0 (Bracco AMT Inc.) to calculate the orbital volume of a patient with orbital trauma preoperatively and postoperatively.11 However, there is no research comparing these different software methods for consistency with or differences from each other. In this study, we used the Ono protocol, developed in the Department of Plastic and Reconstructive Surgery of the Federal University of Parana´.8,12 The authors assert that this method is accurate, reliable, reproducible, cost effective, and relatively fast (the CT volumetric analysis of each orbit takes about 20 minutes). As with every study, there is a constant concern to validate one’s method. As there is no current gold standard method against which to measure one’s results and accomplish this validation, we chose to use a rapid prototyping technology to compare to and to validate the Ono method. This is a relatively new technology, not yet used in other studies for these purposes. Prototyping reconstructs the orbit
with precise anatomical detail, including accurate representation of complex reliefs. If other more accurate methods are available, the use of the dry skull method should be discouraged because the bone has undergone both a dehydration process and loss of periosteum. This shrinking leads to an increase in measured volume and inaccurate results. With the use of prototyping, it was possible to reconstruct the orbit in minute detail and establish precise new analyses of true orbital volume. With respect to measurement of normal values for orbital volume, Deveci et al reported an average volume of 28.41 cm3. Forbes et al reported a mean of 23.9 cm3 for males and 23.63 cm3 for females.6,9 In our study, the average orbital volume measurements obtained by the Ono method of was 20.51 cm3, the average obtained by the SolidWorks program was 20.64 cm3, and the average obtained using the prototype method was 21.81 cm3. These values are slightly less than what is mentioned in the literature. These variations in the absolute values of the orbital volume are due to different methods of measurement used and the type of windowing (for CT scans) used for calculation. Kwon et al in 2009 demonstrated that the orbital volume obtained in the coronal scan was underestimated, as compared with the volumes obtained in axial scans.11 In this project, the 2 planes were included to perform the measurements. The overly high values referenced in the aforementioned studies were probably due to some of the following factors: using the orbits from dry skulls, use of incorrect or imprecise anatomical landmarks, and use of overly thick (1.5 mm) tomographic slices. The method proposed by Ono et al uses finer slices of 0.5 mm thickness to afford a lesser margin for error and subsequent overestimation. Evaluating the absolute values of the orbital volumes for the 3 methods (Ono, SolidWorks, and prototype), it appears that the prototype method is overestimating by about 1 cm3 on average. However, with this limited sample, this is an observation, not a certainty. It may be that, in reality, the orbital volumes obtained by the Ono
TABLE 1. Absolute Orbital Volume Orbital Volume Right, cm3
Orbital Volume Left, cm3
Patient
Ono
SolidWorks
Prototype
Ono
SolidWorks
Prototype
J. S. A. L. R. M.
22.05 22.66 17.88 23.26 19.84 18.36
21.06 23.32 17.03 21.44 20.32 18.09
24.2 24.6 18.9 23.3 22.6 18.2
x 22.35 17.66 23.40 20.06 18.12
x 24.05 19.66 22.56 20.50 18.99
x 24.8 19.3 23.1 22.4 18.5
FIGURE 5. Orbital volume analysis using the 3 methods.
* 2013 Mutaz B. Habal, MD
Copyright © 2013 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
1879
The Journal of Craniofacial Surgery
Roc¸a et al
& Volume 24, Number 6, November 2013
TABLE 3. Comparison Between the Volumes of Right and Left Sides
Method Ono
SolidWorks
Prototype
Side
Average
Minimum
Maximum
Standard Deviation
D E Dif E-D D E Dif E-D D E Dif E-D
20.40 20.32 j0.08 20.04 21.15 1.11 21.52 21.62 0.10
17.88 17.66 j0.31 17.03 18.99 0.18 18.20 18.50 j0.20
23.26 23.40 0.22 23.32 24.05 2.63 24.60 24.80 0.40
2.46 2.53 0.24 2.53 2.10 0.92 2.82 2.65 0.28
P 0.493*
0.043**
0.473*
*Student t test for paired samples, P G 0.05; **Wilcoxon test, nonparametric, P G 0.05.
FIGURE 7. Comparison between the volumes of the left and right sides with the Ono method.
and SolidWorks methods are underestimated. Further study with a larger sample will be required to further evaluate this discrepancy. Of particular note is that the volume obtained by SolidWorks is significantly different when compared to the actual volume of the prototype. As the prototype was generated from this software, it was expected that the values would be similar, which did not occur. The values obtained with the Ono et al method versus those obtained with SolidWorks showed no significant differences. Over the course of this study, the authors developed 2 hypotheses that might explain the higher volumes obtained using the prototype method: either higher volumes calculated using the prototypes may be due to the irregularities of the inner side of the orbit, which are not taken into account by the Ono and SolidWorks methods, both of which use linear slices of the orbits to measure the volumes, or the surface tension of the saline used to fill the orbits cannot be completely broken, and the inserted liquid can exceed the anatomical limits stipulated for measuring orbital volume. Further study is needed to see which, if either, of these hypotheses holds true. The question that remains is: are these small differences in volumes clinically and surgically significant? Few studies were able to measure how the relationship between orbital content and continent is affected by small changes in orbital volume. In 1994, Whitehouse et al showed in a series of cases that an increase in the orbital volume of 1 cm3 generates about 0.77 mm of enophthalmos. Moreover, in 2008, Ahn et al demonstrated that enophthalmos greater than 2 mm typically indicates need for surgical intervention.13,14 In 2010, Yoshida et al utilized the Osirix software v.3.6.16 to assess the relationship between the change in orbital volume
FIGURE 6. Comparison between the volumes of right and left sides.
1880
and the change in ocular projection in patients with zygomatic complex fractures.15 However, the volumes calculated seem to be underestimated. To wit, the volume for the eyeball alone is about 7 cm3.16 Including the extraocular muscles and orbital fat, the volume is 23 to 28 cm3. There may have been some variations or errors when defining the anatomical limits for calculating the orbital volume, or the ROI program (region of interest/computed volume) used in the Osirix software has not correctly measured these volumes. Many studies have tried to establish correlations between the absolute volumes of the right and left orbits of the same patient, without much success.9,17 In this project, it was found that the Ono method determined that the right and left orbits had similar volumes. This finding is in line with the work of Ono et al, 2008, in which it was shown that the orbit of a healthy patient can serve as control for the affected orbit.8,12 This enables a more accurate diagnosis in monitoring patients in the postoperative period, leading to an earlier diagnosis of unwanted changes. The use of rapid prototyping brings greater reliability to the tomographic measurements obtained by the Ono et al method and has been shown to be more reliable than silicone models.10,17 This modeling technology is now increasingly accessible and less expensive. By combining imaging methods with rapid prototyping, craniomaxillofacial surgery enters a new era. In addition to having detailed images of the anatomy, the surgeon may perform craniofacial training using biomedical models to improve physician skills, minimize procedure time, and facilitate better outcomes. More judicious analyses can be made postoperatively and unwanted changes more readily detected.
FIGURE 8. Comparison between the volumes of the left and right sides in SolidWorks method.
* 2013 Mutaz B. Habal, MD
Copyright © 2013 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
The Journal of Craniofacial Surgery
& Volume 24, Number 6, November 2013
5. 6.
7.
8.
9. FIGURE 9. Comparison between the orbital volumes of right and left sides in the prototype.
The tomographic method for the analysis of orbital volume using the Ono protocol yielded consistent but uncorroborated values. This method, combined with rapid prototyping, offered validation of those measurements and greater reliability. Also noted were confirmed similarities in volume for the right and left orbits of individual patients, a long-debated subject. Subsequently, it is reasonable to believe that the tomographic method proposed by Ono can be used in protocols involving diagnosis and therapy of pathologies that affect the orbital region, and the contralateral healthy orbit can serve as an anatomical parameter.
10.
11.
12.
13.
REFERENCES 1. Cooper WC. A method for volume determination of the orbit and its contents by high resolution axial tomography and quantitative digital image analysis. Trans Am Ophthalmol Soc 1985;83:546Y609 2. Charteris DG, Chan CH, Whitehouse RW, et al. Orbital volume measurement in the management of pure blowout fractures of the orbital floor. Br J Ophthalmol 1993;77:100Y102 3. Ellis E, El-Attar A, Moos KF. An analysis of 2,067 cases of zygomatico-orbital fracture. J Oral Maxillofac Surg 1985;43:417Y428 4. Kolk A, Pautke C, Wiener E, et al. A novel high-resolution magnetic resonance imaging microscopy coil as an alternative to the multislice computed tomography in postoperative imaging of orbital fractures and
14.
15.
16. 17.
Comparison of Orbital Volume
computer-based volume measurement. J Oral Maxillofac Surg 2005;63:492Y498 Alexander JC, Anderson JE, Hill JC. The determination of orbital volume. Trans Can Ophthalmolog Soc 1961;24:105Y111 Deveci M, Oztu¨rk S, Sengezer M, et al. Measurement of orbital volume by a 3-dimensional software program: an experimental study. J Oral Maxillofac Surg 2000;58:645Y648 Bite U, Jackson IT, Forbes GS, et al. Orbital volume measurements in enophthalmos using three-dimensional CT imaging. Plast Reconstr Surg 1985;75:502Y508 Ono MC, Eiji S, Freitas RS, et al. Me´todo de comparac¸a˜o do volume orbital por tomografia computadorizada helicoidal. Rev Soc Bras Cir Craniomaxilofac 2008;11:22Y26 Forbes G, Gehring DG, Gorman CA, et al. Volume measurements of normal orbital structures by computed tomographic analysis. AJR Am J Roentgenol 1985;145:149Y154 Regensburg NI, Kok PHB, Zonneveld FW. A new and validated CT-based method for the calculation of orbital soft tissue volumes. Invest Ophthalmol Vis Sci 2008;49:1758Y1762 Kwon J, Barrera JE, Jung T-Y, et al. Measurements of orbital volume change using computed tomography in isolated orbital blowout fractures. Arch Facial Plast Surg 2009;11:395Y398 Ono MC, Eiji S, Freitas RS, et al. Manutenc¸a˜o do volume orbital no po´s-operato´rio tardio de pacientes com fraturas simples do osso zigoma´tico tratados com reduc¸a˜o e fixac¸a˜o incruenta com fio de kirschner. Monografia apresentada ao Servic¸o de Cirurgia Pla´stica e Reparadora da UFPR 2009 Ahn HB, Ryu WY, Yoo KW, et al. Prediction of enophthalmos by computer-based volume measurement of orbital fractures in a Korean population. Ophthal Plast Reconstr Surg 2008;24:36Y39 Whitehouse RW, Batterbury M, Jackson A, et al. Prediction of enophthalmos by computed tomography after ‘blow out’ orbital fracture. Br J Ophthalmol 1994;78:618Y620 Yoshida M, Alonso N, Tonello C, et al. Ana´lise tomogra´fica da relac¸a˜o entre o volume orbita´rio e a projec¸a˜o do globo ocular em fraturas do assoalho orbita´rio. Rev Bras Cir Craniomaxilofac 2010;13:23Y26 Acer N, Sahin B, Ergu¨r H, et al. Stereological estimation of the orbital volume: a criterion standard study. J Craniofac Surg 2009;20:921Y925 Parsons GS, Mathog RH. Orbital wall and volume relationships. Arch Otolaryngol Head Neck Surg 1988;114:743Y747
* 2013 Mutaz B. Habal, MD
Copyright © 2013 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
1881
Comparison of Orbital Volume Obtained by Tomography and Rapid Prototyping Guilherme Berto Roc¸a, MD,* Jose´ Aguiomar Foggiatto, PhD,Þ Maria Cecilia Closs Ono, MD,þ Sergio Eiji Ono, MD,§ and Renato da Silva Freitas, PhD|| Purpose: This study aims to compare orbital volume obtained by helical tomography and rapid prototyping. Methods: The study sample was composed of 6 helical tomography scans. Eleven healthy orbits were identified to have their volumes measured. The volumetric analysis with the helical tomography utilized the same protocol developed by the Plastic Surgery Unit of the Federal University of Parana´. From the CT images, 11 prototypes were created, and their respective volumes were analyzed in 2 ways: using software by SolidWorks and by direct analysis, when the prototype was filled with saline solution. For statistical analysis, the results of the volumes of the 11 orbits were considered independent. Results: The average orbital volume measurements obtained by the method of Ono et al was 20.51 cm3, the average obtained by the SolidWorks program was 20.64 cm3, and the average measured using the prototype method was 21.81 cm3. The 3 methods demonstrated a strong correlation between the measurements. The right and left orbits of each patient had similar volumes. Conclusions: The tomographic method for the analysis of orbital volume using the Ono protocol yielded consistent values, and by combining this method with rapid prototyping, both reliability validations of results were enhanced. Key Words: Orbit, tomography, diagnostic (J Craniofac Surg 2013;24: 1877Y1881)
A
large number of congenital, vascular, endocrinal, tumoral, and traumatic anomalies are able to change the orbital anatomy. The What Is This Box? A QR Code is a matrix barcode readable by QR scanners, mobile phones with cameras, and smartphones. The QR Code links to the online version of the article.
From the *Federal University of Parana´; †Federal University of Technology of Parana´; ‡Plastic Surgery Unit, Federal University of Parana´; §Department of Radiology of DAPI; and ||Plastic Surgery Unit, Federal University of Parana´, Curitiba, Brazil. Received March 18, 2013. Accepted for publication May 5, 2013. Address correspondence and reprint requests to Dr. Guilherme Berto Roc¸a, Plastic Surgery Unit, Federal University of Parana´, Rua General Carneiro, 181VHospital de Clı´nicas Curitiba, CEP 80060-900, Curitiba, Brazil; E-mail: [email protected] This is a Thesis of Master Degree of G.B.R., Post Graduate of SurgeryVFederal University of Parana´. The authors report no conflicts of interest. Copyright * 2013 by Mutaz B. Habal, MD ISSN: 1049-2275 DOI: 10.1097/SCS.0b013e31829a8332
The Journal of Craniofacial Surgery
variation of orbital volume from several pathologies can lead to profound changes in ocular physiology.1 Many diseases that alter the orbit can be treated by surgical procedures to recover its volume. Surgical planning for these patients encounters difficulties due to lack of accuracy and precision inherent with a majority of the evaluation methods.2 Among acquired diseases affecting orbital volume, Graves ophthalmopathy is one of the most frequent. Its association with hyperthyroidism was first described by Robert Graves in 1835. The disease is well known, with a clinical presentation of proptosis caused by an imbalance in the relationship between the orbital content and continent. Among the traumatic causes, facial trauma with fracture of orbital walls commonly alters the orbital volume, with important clinical implications.3 As the orbit is both a functional and aesthetic unit of the face, abnormality of orbital volume is apparent and significant.4 The concept has already been established that restoration of orbital volume can relieve ocular symptoms.5 The methods for evaluation of orbital volume as currently described are difficult to reproduce, and require specific equipment and training. Moreover, there are marked discrepancies in values assigned by different observers.6 Therefore, in clinical practice, the surgeon must rely on his/ her personal observational and surgical skills. A reliable and accurate method could help in understanding the physiopathology and perioperative management of these patients, and more objective volumetric evaluation could help in long-term postoperative follow-up of previously treated patients.7 Undoubtedly, the development of a new radiological method for measuring orbital volumes with easier implementation, improved accuracy, and low cost would greatly benefit patients with orbital fractures and volumetric changes requiring surgical intervention.
PATIENTS AND METHODS This study compared orbital volumes obtained by helical computed tomography and rapid prototyping, using helical CT scans of 6 patients with no evidence of orbital volume abnormality, from a database of the Radiology Department of DAPI (Advanced Diagnostic Imaging), located in Curitiba, Parana´. Sample selection included examination of the data for patients between 20 and 60 years of age (mean age 36 years) with no anomalies reported and no reported medical conditions that would alter the orbital volume, such as sinus disease and retinal disorders. Patients whose reports showed evidence of ocular ophthalmopathy, malformations, and/or tumors were excluded. Six patients with a total of 11 orbits were eventually selected. One patient showed a change in the left orbital floor, and that orbit was excluded from the project. This study was conducted in 3 centers: the Plastic and Reconstructive Surgery Unit of the Federal University of Parana´, the Department of Radiology of DAPI, where Dr. Sergio Ono was responsible for conducting the volumetric analyses in helical CT scans, and the Federal University of Technology of Parana´, at the Prototyping Nucleus (NUFER), under Prof. Dr. Jose´ A. Foggiatto, responsible for the creation and production of the 11 prototypes.
& Volume 24, Number 6, November 2013
Copyright © 2013 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
1877
The Journal of Craniofacial Surgery
Roc¸a et al
& Volume 24, Number 6, November 2013
FIGURE 1. A, B, The Ono method for analyzing orbital volume using a CT scan.
The orbital volume obtained by spiral CT (Ono method) was performed by a single radiologist (Dr. Sergio Ono), in compliance with previously published protocol.8 Using this method, 17 slices of 0.5 mm of thickness were obtained from the CT scans. Then, with the sum of the consecutive cones, the volume of the orbit was calculated (Figs. 1A, B). To analyze the orbital volume in the prototype, 11 models were created and produced by the Stratasys FDM 2000 machine using copolymer ABS (acrylonitrile, butadiene, and styrene) (Fig. 2). The average time to manufacture each orbit was 3 hours. To measure the orbital volume of the prototypes, they were filled with saline solution. The saline was used as it is a noncorrosive material, easy to handle, and offers volume values in cubic centimeters. A drop of detergent was added to the saline solution to break the surface tension and reduce measurement errors. Otherwise, the volume injected into each orbit could exceed the anatomical boundaries used for landmarking (Fig. 3). During the fabrication process of the prototypes using the InVesalius program, it was possible to make a new assessment for orbital volume analysis using the SolidWorks software (Fig. 4). Therefore, 3 methods were available to measure the orbital volume: the Ono method, SolidWorks analysis, and the prototype method (direct measurement with saline). The 3 methods of orbital volumetric measurements were described by mean, median, minimum values, maximum values, and standard deviations. Data were analyzed with Statistica v.8.0 software.
RESULTS After obtaining the volumes of the 11 orbits, the results were tabulated and compared with volumes obtained by imaging methods (Table 1). We tested the null hypothesis that the mean volume is equal to the 3 orbital measurement methods versus the alternative
FIGURE 2. The prototype.
1878
hypothesis that at least one of the methods has a different average than the others. Table 2 and Figure 5 show descriptive statistics of the volume according to the method and the P value of the statistical test. The results showed no significant difference between the volume measured between the Ono method and the SolidWorks method (P = 0.724). There was, however, a significant difference between the Ono and prototype methods (P = 0.001) and between the SolidWorks and prototype methods (P = 0.003) in comparing average orbital volume. For each method, we tested the null hypothesis that the mean volume on the right side is equal to the average volume on the left side versus the alternative hypothesis of different means. The results in Table 3 and in Figures 6, 7, 8, and 9 demonstrated that with the Ono and the prototype methods, the orbital volumes between left and right orbit are not significantly different; however, with the SolidWorks method, this difference was significant.
DISCUSSION Tomography has been the most frequently used method for measuring orbital volume.1,9 Current protocols for the evaluation of orbital volume are not universally accepted. Every radiology and craniomaxillofacial surgery service uses a method and a standard program for orbital volumetric analysis. However, despite these various available methods of measurement, no single method is ideal. Furthermore, most published studies have not compared their results against other measurement methods. Therefore, it is difficult to know how reliable these results are. Several studies have attempted to compare the volumetric analyses obtained by tomography versus direct volumetric measurement of the orbits of skulls. Cooper et al compared the orbital volume obtained with CT against the volume obtained by filling the orbits of skulls with dry sand.1 The discrepancy between the
FIGURE 3. The prototype filled with saline.
* 2013 Mutaz B. Habal, MD
Copyright © 2013 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
The Journal of Craniofacial Surgery
& Volume 24, Number 6, November 2013
Comparison of Orbital Volume
TABLE 2. Orbital Volume Analysis in the Three Methods Method
n
Average
Minimum
Maximum
Standard Deviation
P*
Ono SolidWorks Prototype
11 11 11
20.51 20.64 21.81
17.66 17.03 18.20
23.40 24.05 24.80
2.29 2.16 2.57
0.002 0.002 0.002
*Analysis of variance with repeated measures, P G 0.05.
FIGURE 4. The SolidWorks method for analyzing orbital volume.
results obtained by the direct method and the tomographic method ranged between 0.2% and 4%. Forbes et al also established comparisons between silicone models obtained using dry skulls and examinations of patients without pathologies affecting their orbits. This work was very helpful in establishing normal values for orbital structures, to be used as standards in future studies. The average for bony orbit volume is about 23 cm3 and the average for intraconal fat is 10 cm3.9 The study of Cooper et al and Forbes et al characterized their tomographic methods as reproducible and with good accuracy; however, in analyzing these studies, there is a simple error: none of them used a gold standard method against which to compare the results, so the claim of objective reliability is somewhat compromised. Various kinds of software have been used to analyze the orbital volume. Regensburg et al used the Mimic (Materialise Inc.) program, which calculates the orbital volume in a manner similar to the method proposed by Ono et al.10 Kwon et al used the Vitrea program, version 3.4 (Vital Images Inc.) and Dextroscope version 1.0 (Bracco AMT Inc.) to calculate the orbital volume of a patient with orbital trauma preoperatively and postoperatively.11 However, there is no research comparing these different software methods for consistency with or differences from each other. In this study, we used the Ono protocol, developed in the Department of Plastic and Reconstructive Surgery of the Federal University of Parana´.8,12 The authors assert that this method is accurate, reliable, reproducible, cost effective, and relatively fast (the CT volumetric analysis of each orbit takes about 20 minutes). As with every study, there is a constant concern to validate one’s method. As there is no current gold standard method against which to measure one’s results and accomplish this validation, we chose to use a rapid prototyping technology to compare to and to validate the Ono method. This is a relatively new technology, not yet used in other studies for these purposes. Prototyping reconstructs the orbit
with precise anatomical detail, including accurate representation of complex reliefs. If other more accurate methods are available, the use of the dry skull method should be discouraged because the bone has undergone both a dehydration process and loss of periosteum. This shrinking leads to an increase in measured volume and inaccurate results. With the use of prototyping, it was possible to reconstruct the orbit in minute detail and establish precise new analyses of true orbital volume. With respect to measurement of normal values for orbital volume, Deveci et al reported an average volume of 28.41 cm3. Forbes et al reported a mean of 23.9 cm3 for males and 23.63 cm3 for females.6,9 In our study, the average orbital volume measurements obtained by the Ono method of was 20.51 cm3, the average obtained by the SolidWorks program was 20.64 cm3, and the average obtained using the prototype method was 21.81 cm3. These values are slightly less than what is mentioned in the literature. These variations in the absolute values of the orbital volume are due to different methods of measurement used and the type of windowing (for CT scans) used for calculation. Kwon et al in 2009 demonstrated that the orbital volume obtained in the coronal scan was underestimated, as compared with the volumes obtained in axial scans.11 In this project, the 2 planes were included to perform the measurements. The overly high values referenced in the aforementioned studies were probably due to some of the following factors: using the orbits from dry skulls, use of incorrect or imprecise anatomical landmarks, and use of overly thick (1.5 mm) tomographic slices. The method proposed by Ono et al uses finer slices of 0.5 mm thickness to afford a lesser margin for error and subsequent overestimation. Evaluating the absolute values of the orbital volumes for the 3 methods (Ono, SolidWorks, and prototype), it appears that the prototype method is overestimating by about 1 cm3 on average. However, with this limited sample, this is an observation, not a certainty. It may be that, in reality, the orbital volumes obtained by the Ono
TABLE 1. Absolute Orbital Volume Orbital Volume Right, cm3
Orbital Volume Left, cm3
Patient
Ono
SolidWorks
Prototype
Ono
SolidWorks
Prototype
J. S. A. L. R. M.
22.05 22.66 17.88 23.26 19.84 18.36
21.06 23.32 17.03 21.44 20.32 18.09
24.2 24.6 18.9 23.3 22.6 18.2
x 22.35 17.66 23.40 20.06 18.12
x 24.05 19.66 22.56 20.50 18.99
x 24.8 19.3 23.1 22.4 18.5
FIGURE 5. Orbital volume analysis using the 3 methods.
* 2013 Mutaz B. Habal, MD
Copyright © 2013 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
1879
The Journal of Craniofacial Surgery
Roc¸a et al
& Volume 24, Number 6, November 2013
TABLE 3. Comparison Between the Volumes of Right and Left Sides
Method Ono
SolidWorks
Prototype
Side
Average
Minimum
Maximum
Standard Deviation
D E Dif E-D D E Dif E-D D E Dif E-D
20.40 20.32 j0.08 20.04 21.15 1.11 21.52 21.62 0.10
17.88 17.66 j0.31 17.03 18.99 0.18 18.20 18.50 j0.20
23.26 23.40 0.22 23.32 24.05 2.63 24.60 24.80 0.40
2.46 2.53 0.24 2.53 2.10 0.92 2.82 2.65 0.28
P 0.493*
0.043**
0.473*
*Student t test for paired samples, P G 0.05; **Wilcoxon test, nonparametric, P G 0.05.
FIGURE 7. Comparison between the volumes of the left and right sides with the Ono method.
and SolidWorks methods are underestimated. Further study with a larger sample will be required to further evaluate this discrepancy. Of particular note is that the volume obtained by SolidWorks is significantly different when compared to the actual volume of the prototype. As the prototype was generated from this software, it was expected that the values would be similar, which did not occur. The values obtained with the Ono et al method versus those obtained with SolidWorks showed no significant differences. Over the course of this study, the authors developed 2 hypotheses that might explain the higher volumes obtained using the prototype method: either higher volumes calculated using the prototypes may be due to the irregularities of the inner side of the orbit, which are not taken into account by the Ono and SolidWorks methods, both of which use linear slices of the orbits to measure the volumes, or the surface tension of the saline used to fill the orbits cannot be completely broken, and the inserted liquid can exceed the anatomical limits stipulated for measuring orbital volume. Further study is needed to see which, if either, of these hypotheses holds true. The question that remains is: are these small differences in volumes clinically and surgically significant? Few studies were able to measure how the relationship between orbital content and continent is affected by small changes in orbital volume. In 1994, Whitehouse et al showed in a series of cases that an increase in the orbital volume of 1 cm3 generates about 0.77 mm of enophthalmos. Moreover, in 2008, Ahn et al demonstrated that enophthalmos greater than 2 mm typically indicates need for surgical intervention.13,14 In 2010, Yoshida et al utilized the Osirix software v.3.6.16 to assess the relationship between the change in orbital volume
FIGURE 6. Comparison between the volumes of right and left sides.
1880
and the change in ocular projection in patients with zygomatic complex fractures.15 However, the volumes calculated seem to be underestimated. To wit, the volume for the eyeball alone is about 7 cm3.16 Including the extraocular muscles and orbital fat, the volume is 23 to 28 cm3. There may have been some variations or errors when defining the anatomical limits for calculating the orbital volume, or the ROI program (region of interest/computed volume) used in the Osirix software has not correctly measured these volumes. Many studies have tried to establish correlations between the absolute volumes of the right and left orbits of the same patient, without much success.9,17 In this project, it was found that the Ono method determined that the right and left orbits had similar volumes. This finding is in line with the work of Ono et al, 2008, in which it was shown that the orbit of a healthy patient can serve as control for the affected orbit.8,12 This enables a more accurate diagnosis in monitoring patients in the postoperative period, leading to an earlier diagnosis of unwanted changes. The use of rapid prototyping brings greater reliability to the tomographic measurements obtained by the Ono et al method and has been shown to be more reliable than silicone models.10,17 This modeling technology is now increasingly accessible and less expensive. By combining imaging methods with rapid prototyping, craniomaxillofacial surgery enters a new era. In addition to having detailed images of the anatomy, the surgeon may perform craniofacial training using biomedical models to improve physician skills, minimize procedure time, and facilitate better outcomes. More judicious analyses can be made postoperatively and unwanted changes more readily detected.
FIGURE 8. Comparison between the volumes of the left and right sides in SolidWorks method.
* 2013 Mutaz B. Habal, MD
Copyright © 2013 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
The Journal of Craniofacial Surgery
& Volume 24, Number 6, November 2013
5. 6.
7.
8.
9. FIGURE 9. Comparison between the orbital volumes of right and left sides in the prototype.
The tomographic method for the analysis of orbital volume using the Ono protocol yielded consistent but uncorroborated values. This method, combined with rapid prototyping, offered validation of those measurements and greater reliability. Also noted were confirmed similarities in volume for the right and left orbits of individual patients, a long-debated subject. Subsequently, it is reasonable to believe that the tomographic method proposed by Ono can be used in protocols involving diagnosis and therapy of pathologies that affect the orbital region, and the contralateral healthy orbit can serve as an anatomical parameter.
10.
11.
12.
13.
REFERENCES 1. Cooper WC. A method for volume determination of the orbit and its contents by high resolution axial tomography and quantitative digital image analysis. Trans Am Ophthalmol Soc 1985;83:546Y609 2. Charteris DG, Chan CH, Whitehouse RW, et al. Orbital volume measurement in the management of pure blowout fractures of the orbital floor. Br J Ophthalmol 1993;77:100Y102 3. Ellis E, El-Attar A, Moos KF. An analysis of 2,067 cases of zygomatico-orbital fracture. J Oral Maxillofac Surg 1985;43:417Y428 4. Kolk A, Pautke C, Wiener E, et al. A novel high-resolution magnetic resonance imaging microscopy coil as an alternative to the multislice computed tomography in postoperative imaging of orbital fractures and
14.
15.
16. 17.
Comparison of Orbital Volume
computer-based volume measurement. J Oral Maxillofac Surg 2005;63:492Y498 Alexander JC, Anderson JE, Hill JC. The determination of orbital volume. Trans Can Ophthalmolog Soc 1961;24:105Y111 Deveci M, Oztu¨rk S, Sengezer M, et al. Measurement of orbital volume by a 3-dimensional software program: an experimental study. J Oral Maxillofac Surg 2000;58:645Y648 Bite U, Jackson IT, Forbes GS, et al. Orbital volume measurements in enophthalmos using three-dimensional CT imaging. Plast Reconstr Surg 1985;75:502Y508 Ono MC, Eiji S, Freitas RS, et al. Me´todo de comparac¸a˜o do volume orbital por tomografia computadorizada helicoidal. Rev Soc Bras Cir Craniomaxilofac 2008;11:22Y26 Forbes G, Gehring DG, Gorman CA, et al. Volume measurements of normal orbital structures by computed tomographic analysis. AJR Am J Roentgenol 1985;145:149Y154 Regensburg NI, Kok PHB, Zonneveld FW. A new and validated CT-based method for the calculation of orbital soft tissue volumes. Invest Ophthalmol Vis Sci 2008;49:1758Y1762 Kwon J, Barrera JE, Jung T-Y, et al. Measurements of orbital volume change using computed tomography in isolated orbital blowout fractures. Arch Facial Plast Surg 2009;11:395Y398 Ono MC, Eiji S, Freitas RS, et al. Manutenc¸a˜o do volume orbital no po´s-operato´rio tardio de pacientes com fraturas simples do osso zigoma´tico tratados com reduc¸a˜o e fixac¸a˜o incruenta com fio de kirschner. Monografia apresentada ao Servic¸o de Cirurgia Pla´stica e Reparadora da UFPR 2009 Ahn HB, Ryu WY, Yoo KW, et al. Prediction of enophthalmos by computer-based volume measurement of orbital fractures in a Korean population. Ophthal Plast Reconstr Surg 2008;24:36Y39 Whitehouse RW, Batterbury M, Jackson A, et al. Prediction of enophthalmos by computed tomography after ‘blow out’ orbital fracture. Br J Ophthalmol 1994;78:618Y620 Yoshida M, Alonso N, Tonello C, et al. Ana´lise tomogra´fica da relac¸a˜o entre o volume orbita´rio e a projec¸a˜o do globo ocular em fraturas do assoalho orbita´rio. Rev Bras Cir Craniomaxilofac 2010;13:23Y26 Acer N, Sahin B, Ergu¨r H, et al. Stereological estimation of the orbital volume: a criterion standard study. J Craniofac Surg 2009;20:921Y925 Parsons GS, Mathog RH. Orbital wall and volume relationships. Arch Otolaryngol Head Neck Surg 1988;114:743Y747
* 2013 Mutaz B. Habal, MD
Copyright © 2013 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
1881
