Three-dimensional study of the orbit-related structures according to sex, age and skeletal deformities

Objective: This study aimed to evaluate the relations between orbit-related structures and sex, age and skeletal deformities using cone-beam computed tomography (CBCT). Methods: This retrospective study evaluated 216 consecutive CBCT scans of patients, who were divided according to: sex (male, n=105; female, n=111), age (A1: 1832 years, n=71; A2: 33-47 years, n=78; A3: 48-62 years, n=67), and skeletal deformities (Class I, n=70; Class II, n=75; Class III, n=71). The supraorbital foramen (SOF) location, volume of orbit, optic canal (OC) and infraorbital canal (IOC) were evaluated. Results were analyzed using the Gamma model test. The Tukey-Kramer post-hoc test was used to compare the variables with three factors (p<0.05). Results: The IOC volume showed higher values for male, A3 and class I patients. The SOF location and the orbital volume also showed higher values for male patients. Regarding the volume of CO, it showed higher values for male and class I patients. Conclusions: According to our results, sex has been shown to have a significant influence on orbit-related structures. Age and skeletal deformities also influenced the volume of IOC and OC. These results eventually help the clinical practice, being useful for orbital reconstruction surgeries, anthropological studies, gender identification and identification of susceptibility to pathological conditions related to sexual dimorphism.


Introduction
The orbit is formed by the maxilla, frontal ethmoid, lacrimal, zygoma, sphenoid and palatine (Oppenheimer, Monson & Buchman, 2013). These are distributed along its walls. Its upper part consists of a part of the frontal bone and the lesser wing sphenoid; the inferior by the orbital plate of the maxilla, the orbital process of the zygoma and the orbital process of the palatine bone (Grob, Yonkers & Tao, 2017). Its medial wall is composed of the frontal process of the maxilla, orbital lamina of the ethmoid, orbital surface of the lacrimal bone, a part of the sphenoid bone. Its lateral wall encompasses the orbital process of zygoma and the orbital surface of the greater wing of the sphenoid (Norton, 2007;Hiatt & Gartner, 2001).
Visual impairment, hypoesthesia of the infraorbital nerve, enophthalmia, irritating and persistent diplopia are considered complications of orbital fractures that are not managed properly (Manana et al., 2017). To avoid unwanted clinical results, orbital reconstruction surgery requires a complete assessment of orbital defects and the precise restoration of orbital dimensions and their surrounding structures (Yang & Liao, 2019). Therefore, a detailed preoperative radiograph evaluation and knowledge of these anatomical structures are necessary to obtain ideal results (Akdemir, Tekdemir & Altin, 2004).
For this reason, knowledge of the orbital volume and the relationship with clinical variables such as age and sex are important in the surgical decision-making process (Andrades et al., 2018). Furthermore, Friedrich et al. (2016) concluded that orbital volume measurements should be included in the follow-up control with special consideration of age-and genderdependent changes in this parameter in orbital reconstructive surgeries. To the best of our knowledge, there are no studies that have evaluated the orbital volume and the structures that surround the orbit, such as the optic canal (OC), the infraorbital canal (IOC), the supraorbital foramen (SOF), and their relationship between sex, age and skeletal deformities. Thus, the aim of this study was to study the relations between orbit-related structures (orbital volume, OC volume, IOC volume, SOF location) and sex, age and skeletal deformities using cone-beam computed tomography (CBCT). The null hypothesis was that the orbit-related structures evaluated do not change according to sex, age and skeletal deformities. Patients of both sexes, >18 years old and diagnosed with skeletal deformities (class I, II and III) were included in this study. Skeletal deformities were classified as class I (0º< ANB <4º), class II (ANB ≥ 4º), and class III (ANB < 0º) (Steiner, 1953). The exclusion criteria adopted in this study were: patients who underwent orthognathic surgery or any surgery on the middle third of the face, which could compromise the analysis of the structures of interest; patients who have craniofacial syndromes and congenital craniofacial anomalies.
Consecutive CBCT scans, which were requested for orthodontic or orthognathic surgery diagnosis and treatment planning, of 216 patients were obtained. All CBCT scans were carried out by the same specialist in oral and maxillofacial radiology and acquired with the i-CAT Next Generation ® equipment (Imaging Sciences International, Hatfield, PA, USA).
Volumes were reconstructed with an isometric voxel size of acquisition of 0.30 mm, FOV (Field of View) of 17 x 23 cm, tube tension of 120 kVp, and tube current of 3-8 mA. All CBCT scans were performed according to a strictly standardized scanning protocol; the patients were instructed to remain seated, adopt a natural head position, breathe lightly with their tongues and lips at rest, and stabilize their heads on a head and chin support.
The CBCTs were transferred into the Dolphin Imaging software version 11.95 (Dolphin Imaging & Management Solutions, Chatsworth, CA, USA) in DICOM (Digital Imaging and Communication in Medicine) format. The Dolphin software contains a tool for semi-automatic segmentation of the airways and its volume measurement, which is called sinus/airway. The sinus/airway tool was used to analyze the volume of the orbit, OC and IOC. This tool requires the user to define an area of interest by using some steps. First, the user must define, manually, the structure boundaries in the sagittal, axial and coronal reconstructions. Second, the user must place, manually, seed points in the destination compartment of the region of interest. The target volume will be filled automatically from those seed points; and all areas with similar grayscale intensity will also be Research, Society andDevelopment, v. 10, n. 11, e77101119381, 2021 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v10i11.19381 4 selected, depending on the Hounsfield unit. Consequently, the threshold value must be determined. This limit defines a density range which will be included in the measured volume (de Water, Saridin, Bouw, Murawska & Koudstaal, 2014). The Hounsfield scale was adjusted and standardized, using the value of 400 HU for orbital volume (Nout et al., 2012), and the value of 70 HU for the OC and IOC volume.
In our study, the limits of each structure were standardized. For the orbital measurement, upper, lower, lateral, and medial limits are determined by the bone walls of the orbit. The anterior orbital margin was defined by a line between the most anterior point of the lateral bone limit and the most anterior point of the anterior lacrimal crest. The posterior limit is defined by the most anterior portion of the optical canal (Andrades et al., 2018;Friedrich, Bruhn & Lohse, 2016). For the OC measurement, the anterior and posterior bone limits of the canal were traced, taking into account the bone point defined more anteriorly and posteriorly in the minor wing of the sphenoid bone (Friedrich et al., 2016). For the IOC measurement, the anterior bone limits begin in the infraorbital foramen and the posterior margin of the IOC covered by the bone of the orbital floor (Fontolliet, Bornstein & Von Arx, 2019) (Figure 1).  The images obtained were analyzed by one specialist in oral radiology, with experience in the analysis of tomographic images and the manipulation of image software. For the calibration process, 20% of CBCT scans were selected randomly and measurements were replicated twice with an interval of 15 days. Intra-examiner agreement was assessed by intraclass correlation coefficient (ICC). The generalized linear models (Gamma model) were used to evaluate the outcome variables (orbital volume, OC volume, IOC volume and SOF location) with sex, age and skeletal deformities. The Tukey-Kramer post-hoc test was used to compare the variables with three factors (age and skeletal deformities). All statistical tests were analyzed with R 3.1 software for Windows (R-project for statistical computing) at a 5% level of significance.

Results
The ICC value was good for all measurements, ranging from 0.732 to 0.941 (Koo & Li, 2016). The 216 scans were composed of 105 male and 111 female patients. Seventy-one individuals were between 18-32 years old; 78 were between 33-47 and 67 between 48-62. Regarding skeletal deformities, 70 individuals were classified as class I, 75 class II and 71 class III.
All the outcome variables (orbital volume, OC volume, IOC volume and SOF location) did not show statistically significant difference in terms of laterality (p>0.05). Therefore, these variables were not analyzed separately. The IOC volume displayed statistically significant difference between sex, age, and skeletal deformities. Male patients showed a larger volume when compared to female. The older group (A3) also exhibited a larger volume than the younger groups (A1 and A2) ( Table 1).
In addition, class I patients presented a statistically larger volume than class II and III (Table 2). Significant statistical differences were noted between sex in SOF location and orbit volume, with higher values for male patients. Regarding the OC volume, significant statistical differences were observed between sex and skeletal deformities.
Patients diagnosed with class I presented a larger volume than class II and class III (Table 2).

Discussion
All the outcome variables of our study showed statistically significant differences between sex. This can be ascribed to the fact that the analyzed structures present sexual dimorphism, which is in agreement with previous studies (Friedrich et al., 2016;Graillon, Boulze, Adalian, Loundou & Guyot, 2017;Erkoç, Öztoprak, Gümüş & Okur, 2015). According to Erkoç et al. (2015), the axial skeleton is larger in men than in women, which also affects the orbit and the surrounding structures. Thus, this can be useful for orbital reconstruction surgeries, anthropological studies and gender identification. Furthermore, susceptibility to pathological conditions has been found in women as a result of these size differences, such as the risk of optic nerve compression and myopia, especially in patients with skeletal deformities (Erkoç et al., 2015).
Identifying and preserving supraorbital and infraorbital foramina are important objectives in surgical and local anesthetic procedures (Aziz, Marchena, & Puran, 2000). In our study, we analyzed the IOC volume, which was higher in the older group (48-62 years) when compared to other age groups. Although no studies had analyzed the IOC volume per se, our results are in line with Fontolliet et al. (2019), which used linear measures to estimate the IOC. They reported that the length and height of the IOC were significantly greater in patients older than 61 years (Fontolliet et al., 2019). Accordingly, the increase in the IOC volume with age can be attributed to the posterior displacement of the maxilla and the lateral and inferior orbital border, which occurs in facial skeletal aging (Lambros, 2007;Lim, Min, Lee & Hong, 2016). In addition, patients in class I had higher IOC volume values than those in classes II and III, showing that in addition to age and sex influence the anatomy of the IOC, skeletal deformities should be considered.
Leaving vital structures such as the optic nerve intact is a challenge in the restoration of complex anatomical structures in craniofacial reconstructive surgery (Kim, Jung, Kim, Lee & Kim, 2013). In our study, we analyzed the volume of the OC in patients with different skeletal deformities, which makes our study unprecedented and may contribute to the characterization of this structure. Our results show that OC volume presents statistical differences according to skeletal deformities, being higher in class I patients than in class II and III patients.
The orbital volume in our study did not change significantly with age. Diverging from previous studies, which reported that the orbital volume increases with age in a continuous way (Ugradar & Lambros, 2019;Friedrich et al., 2016). The results of our study suggest that the bone of the orbit is dynamic throughout adult life and the clinical signs of periorbital aging maybe not due to orbital bone changes. The same explanation can be attributed to the OC volume in our study, suggesting that just as orbital bones are dynamic throughout adulthood.
In our study, the SOF location showed a non-significant increase in patients aged 48-62 years. Lim et al. (2016) observed that older patients displayed the supraorbital foramen significantly more distant from the midline, which indicates lateral translation of the frontal and maxillary bones with aging.
In our study, the measurement of the orbital volume was carried out due the fact that craniofacial reconstructive surgery has the fundamental principle of restoring facial symmetry, reconstructing the defective orbital wall, and restoring the appropriate orbital volume to mitigate restricted eye movements (Kim et al., 2013). However, differences in the orbital volume in relation to other studies have been noticed, which may be due, at least in part, to the different image sources used. Most studies were performed with computed tomography (Andrades et al., 2018;Diaconu et al., 2017). Currently, CBCT has been providing useful data for defining orbital pathologies, calculating orbital reconstructions, or anthropological studies (Friedrich et al., 2016).
Moreover, the use of CBCT is considered adequate for the evaluation of the orbital volume, as long as reference points are defined to a closed surface for the calculations (Friedrich et al., 2016). For this reason, CBCTs were used in our study to assess the orbit and its adjacent structures, taking into account the standardization of bone limitations of all analyzed structures. This may also be responsible for the differences observed in the values of the orbital volume and other structures among the studies.
It is worth mentioning that our sample size is remarkable and very standardized, since all CBCT scans were cautiously conducted by a single radiologist with a standardized scanning protocol. However, the comparison between studies that evaluated orbit-related structures has been proven to be difficult, because there is no consensus in the literature with the methodologies.
Previous studies used different delimitations of orbital borders and surrounding structures for measurement, which made it difficult to compare the volumetric data. Furthermore, the literature regarding the orbit and the surrounding structures is still scarce, especially with respect to skeletal deformities. The results of this study bring a new perspective for the anatomy of this region.

Conclusion
According to our results, sex has been shown to have a significant influence on orbit-related structures. Age and skeletal deformities also influenced the volume of IOC and OC. These results eventually help the clinical pratice, since this information could be useful for orbital reconstruction surgeries, anthropological studies, gender identification and identification of susceptibility to pathological conditions related to sexual dimorphism.