Three-Dimensional Morphometric Evaluation of the Orbital Aperture in Multislice Computed Tomography: Anatomical Classification and Deep Learning-Based Sex Estimation

Three-Dimensional Morphometric Evaluation of the Orbital Aperture in Multislice Computed Tomography: Anatomical Classification and Deep Learning-Based Sex Estimation | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Three-Dimensional Morphometric Evaluation of the Orbital Aperture in Multislice Computed Tomography: Anatomical Classification and Deep Learning-Based Sex Estimation Sibel Ateşoğlu Karabaş, Mehlika Küçük Yanar, Duygu Akın Saygın, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6496345/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract The orbital aperture (OA) is an important anatomical structure that forms the entrance of the orbit and has connections with intracranial structures. This structure has a critical importance for clinicians as it contains reference points in surgical approaches and requires attention in plastic reconstructive surgery. The OA is also one of the craniofacial variables used for sex estimation in anthropology and forensic medicine. The aim of this study was to evaluate the OA morphologically and to determine its role in sex estimation by deep learning method. Three-dimensional (3D) images of 100 adult individuals (48 males, 52 females) on Multislice Computed Tomography (MSCT) were used in the study. A classification for the OA shape was created by classifying the right and left side OA images based on observation by 3 researchers. In this study, sex estimation was performed using convolutional neural network (CNN) models to extract deep features found in OA coronal and MSCT images. Gender prediction percentage was calculated using deep learning method over the OA images registered to the algorithm one by one. For OA morphology, 5 types were identified: square, trapezoid, round, oval, round-trapezoid. The percentage of gender prediction, with the deep learning method, in coronal slices and 3D images was found to be 65.6% and 73.4%, respectively. The most common OA shape was round-trapezoid (74%) in both coronal and 3D images in males, while it was round type (39%) in females. In CNN modeling, incorporating OA types along with side information led to a decrease in gender estimation accuracy. This highlighted the importance of considering the morphological variations of OA and their distribution across sides. Interestingly, when side information is excluded, gender prediction accuracy can exceed 80% in both coronal and 3D images. Health sciences/Anatomy Physical sciences/Mathematics and computing Orbit Multislice Computed Tomography Deep Learning Sex estimation Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction The orbital aperture (OA) refers to the opening of the orbit and is an anatomical landmark in the skull. The OA is formed by the frontal, maxillary and zygomatic bones. It has four edges called supraorbital, infraorbital, medial and lateral margins [ 1 ]. These thickened margins formed by bony structures surround the OA and protect the eye against impacts. The size and shape of the OA varies between populations and genders. These differences are thought to be mainly due to changes in the slope of the superior and inferior margins. It is known that the medial margin is flat and narrow, and the superior margin is more rounded in Asian populations, whereas the superior and inferior margins are more horizontal and parallel in European populations [ 2 ]. The OA is very important for anthropologists and maxillofacial surgeons, and it shows differences between populations. With the developing technology, there is a need to renew the gross anatomical classifications of OA or to evaluate the reliability of existing classifications [ 3 , 4 ]. A good knowledge of the morphology of the OA, which differs between societies, can be used as an important parameter in determining age, sex and race in forensic medicine [ 5 ]. Sex estimation is an interdisciplinary field requiring expertise used in many fields, especially in forensic medicine and anthropology [ 6 ]. According to the skeletal parts, the most dimorphic region after the pelvis is considered to be the skull with a success rate of 90% [ 7 , 8 ]. Today, with the widespread use of computer-aided analysis systems, the focus is on systems that can support experts in the decision-making process. For this, machine learning algorithms are often preferred due to their low cost, more reliable and faster results. Machine learning algorithms also increase the predictive power thanks to their ability to combine and manage large and complex data [ 9 ]. Unlike traditional machine learning, image, video, audio and text symbols are processed. Today, its use in the field of medicine is becoming increasingly widespread due to its high success rates in gender and age prediction [ 10 ]. In the light of the above information, we set two aims and objectives in our study. The first was to create OA typing in coronal sections and 3D images. Secondly, we aimed to determine the role of deep learning method in gender determination using coronal sections and 3D OA images. Methods Study Population The study was approved by the decision of Kahramanmaraş Sütçü İmam University Medical Research Ethics Committee No. 2023/21 & 07. In the study, Multislice Computed Tomography (MSCT) images of patients who applied to Necmettin Erbakan University Faculty of Medicine, Department of Radiology between 2020 and 2023 were retrospectively analysed. A total of 100 TMSCT images of 100 individuals (52 females, 48 males) between the ages of 18–65 years obtained as a result of exclusion criteria (deformity, pathological lesion, trauma or surgical operation history in the orbital, intraorbital and periorbital region, incompletely visualized artefacts and images with low image quality that do not meet the research requirements) were evaluated. In addition, individuals aged 18–65 years were categorized into four groups as 18–24, 25–34, 35–44 and 45–65. All participants were given the necessary information about the purpose and importance of the study and were informed that MSCT images would be used. Written consent was obtained from all participants. The study was also conducted in accordance with the principles stated in the Declaration of Helsinki. Study Design Temporal 0.625 mm thick temporal MSCT images of the individuals whose heads were taken in the supine position by placing the head in the neutral position with a 256-slice, double-tube MSCT Somatom Drive (Siemens Healthineers, Germany) device was obtained in the PACS system. The imaging parameters of these images were: kV, 80 and 140; mA, 89 and 178; rotation time, 0.28 s; collimation, 256x0.625; FOV, 220 mm. Images were analysed in coronal section using the PACS measurement module. Temporal TMSCT images were saved in Digital Imaging and Communications in Medicine (DICOM) format and 3D images were obtained using RadiAnt DICOM Viewer, an open-source software platform. The images obtained were created by an investigator with 10 years of radiological experience, considering the exclusion criteria. Morphological Evaluation of OA The shape evaluation of the OA obtained with coronal section and 3D images was performed by 3 different researchers at different times, and then the results were compared, and the disagreeing opinions were re-examined by 3 researchers and a consensus was reached. In the morphological evaluation of OA, a new typing was created by modifying the Szilvassy classification. OA shape was classified into 5 types as type 1: square, type 2: trapezoid, type 3: round, type 4: oval and type 5: round-trapezoid (Fig. 1 – 2 ). Deep Learning Attributes are vector quantities with a high level of representation that are used in the interpretation of a data. In order to obtain these values, edges, corners, colors, morphological features, etc. are used in image data. Obtaining features in this way is traditional feature extraction. These attributes are calculated and classified with new generation approaches known as convolutional neural networks (CNN), which consist of layers such as input, convolution, normalization, activation, pooling, fully connected layer and output (classification) layer. After the image data is directly given to the CNN model input layer, there is no need for an external feature extraction method. This task is already performed by the filters in the convolution layers of the CNN model. The features calculated up to the fully connected layer are then classified in the classifier layer using the SoftMAX function. CNN models using these features are widely used due to their high performance in the classification task [ 11 ]. EfficientNet CNN model is one of the new generation convolutional neural network architectures that provides cost advantage as it can control the computational cost more efficiently while increasing the depth and width of the network. The depth, width and resolution values of the network are scaled uniformly, and the instability and inefficiency caused by random scaling are eliminated. This is achieved by using the compound coefficient technique. In this study, the second generation EfficientNet model, EfficientNetv2b0 architecture was used. The number of parameters, which was 5.3M in EfficientNet b0 version, increased up to 7.2M in v2 b0 version. The features were obtained from the fully connected layer of the model. These features were classified with the Support Vector Machine (SVM) classifier using linear and polynomial (quadratic and cubic) Kernel function 44. SVM classifier is a widely used and successful classifier method [ 12 ]. The method searches for a hyperplane that can best separate the data. SVM classifer can also work if the number of classes is more than two. SVM classifer using linear, quadratic, and cubic kernel functions was used in the study [ 13 ]. The algorithm attempts to minimise a scalar objective function, f(x), for values of ‘x’ limited to a finite space. This objective function can be deterministic or stochastic, i.e. it can produce different results when evaluated at the same point x [ 13 ]. In this study, Expected Improvement (EI) was used to model the criteria for selecting the next hyperparameter. Eq. 2 models this function, where, denotes the threshold value of the objective function and ‘’y‘’ represents the current value of the objective function. The objective is to maximise the value of EI using this equation, taking into account the set of hyperparameters. The equation for the dual version of the general SVM method is given in Eq. (1) below. In the study, data sets were obtained by recording coronal slice and 3D images for deep learning (Fig. 1 ). Images of the OA were given to the input layer of the EfficientNetv2b0 model. Image features were then extracted from the fully connected layer (dense) of the pre-trained CNN model with 1000 features for each image data. Thus, 987x1000 features were obtained for a total of 987 data, 496 data in each class. These features were trained separately using three different Kernel functions of the SVM classifier. Bayesian optimisation was then used to determine the ideal values of the classifier hyperparameters. Classifier training was completed according to the optimized hyperparameters, and validation was performed by cross-validation. A 10-fold cross-validation approach was used to measure the performance of the trained classifier. Statistical Analysis The minimum sample size was determined as 70 (α [significance level] = 0.05; 1-β [power] = 0.95) by power analysis (G*power 3.1.9.4) in accordance with the data obtained from the literature. Statistical analyses were performed with IBM Statistical Package for the Social Sciences (SPSS) Statistics 21.0 (Chicago Illinois). Results Of the 100 individuals aged between 18 and 65 years included in our study, 48 were male and 52 were female. The mean age was 38.34 ± 13 years (male, 36.25 ± 11.56; female, 40.27 ± 14.04) and the data of the individuals were evaluated in 4 groups as 18–24 years (20 individuals, mean age 20.80 ± 1.85), 25–34 years (22 individuals, mean age 29.91 ± 2.61), 35–44 years (24 individuals, mean age 40.08 ± 2.91) and 45–65 years (34 individuals, mean age 52.88 ± 6.43). The difference between genders in age groups was statistically significant (χ2: 9.213; p = 0.027) (Table 1 ). Table 1 Distribution by age groups and sex. Total (n = 100) Male (n = 48) Female (n = 52) Ages n % n % n % p 18–24 20 20 9 18.8 11 21.2 25–34 22 22 13 27.1 9 17.3 0.027* 35–44 24 24 16 33.3 8 15.4 45–65 34 34 10 20.8 24 46.2 *p < 0,05, Chi-square test. In observation-based classification, 5 different OA types were determined (Fig. 2 ). Classification was performed separately for each side. In coronal section images, type 5 (round-trapezoid) was observed most frequently on both sides in both sexes. In 3D images, it was observed that the OA shape was mostly type 5 (round-trapezoidal) in males and type 3 (round) in females, while type 3 (round) was the most common type in the right OA and type 5 (round-trapezoidal) was the most common type in the left OA (Table 2 ). There was a statistically significant difference between genders in left coronal OA types (p = 0.024). Table 2 Male and female AO shapes in coronal section and 3D images. Total (n = 100) Male (n = 48) Female (n = 52) Types n % n % n % p Coronal Right Type 1 1 1 1 2.1 - - Type 2 16 16 10 20.8 6 11.5 Type 3 11 11 2 4.2 9 17.3 0.149 Type 4 3 3 2 4.2 1 1.9 Type 5 69 69 33 68.8 36 69.2 Left Type 1 3 3 3 6.3 - - Type 2 12 12 6 12.5 6 11.5 Type 3 8 8 - - 8 15.4 0.024* Type 4 3 3 1 2.1 2 3.8 Type 5 74 74 38 79.2 36 69.2 3D Right Type 1 13 13 7 14.6 6 11.5 Type 2 15 15 8 16.7 7 13.5 Type 3 35 35 15 31.3 20 38.5 0.886 Type 4 3 3 2 4.2 1 1.9 Type 5 34 34 16 33.3 18 34.6 Left Type 1 14 14 7 14.6 7 13.5 Type 2 17 17 10 20.8 7 13.5 Type 3 30 30 10 20.8 20 38.5 0.411 Type 4 4 4 2 4.2 2 3.8 Type 5 35 35 19 39.6 16 30.8 When analysed according to age groups, type 5 was the most common type in coronal measurements in all age groups, while it was found to vary between age groups in 3D images. In the age range of 18–24 years, type 3 (round) was the most common type of OA on both sides, while type 5 (round-trapezoid) was observed in the age range of 44–65 years. There was a statistically significant difference between left-sided OA types and age groups (p < 0.05) (Table 3 ). Table 3 Orbital types according to age groups in coronal sections and 3D images. Age Groups (years) 18–24 25–34 35–44 45–65 Types n % n % n % n % p Coronal Right Type 1 - - 1 4.5 - - - - Type 2 4 20 6 27.3 6 25 - - Type 3 3 15 1 4.5 - - 7 20.6 0.052 Type 4 1 5 1 4.5 - - 1 2.9 Type 5 12 60 13 59.1 18 75 26 76.5 Left Type 1 1 5 1 4.5 1 4.2 - - Type 2 3 15 5 22.7 4 16.7 - - Type 3 3 15 - - - - 5 14.7 0.136 Type 4 1 5 1 4.5 - - 1 2.9 Type 5 12 60 15 68.2 19 79.2 28 82.4 3D Right Type 1 2 10 4 18.2 6 25 1 2.9 Type 2 1 5 2 9.1 6 25 6 17.6 Type 3 11 55 8 36.4 5 20.8 11 32.4 0.85 Type 4 - - - - 2 8.3 1 2.9 Type 5 6 30 8 36.4 5 20.8 15 44.1 Left Type 1 1 5 6 27.3 5 20.8 2 5.9 Type 2 4 20 2 9.1 6 25 5 14.7 Type 3 11 55 5 22.7 5 20.8 9 26.5 0.011* Type 4 - - 1 4.5 3 12.5 - - Type 5 4 20 8 36.4 5 20.8 18 52.9 Sex estimation performance Coronal section and 3D images were recorded differently and data sets were created. With deep learning, gender detection in coronal images was 65.6%, while gender detection in 3D images was 73.4%. Deep learning confision matrix and training graphs are given in Fig. s 3 & 4. Discussion Sex estimation is one of the most important milestones in identification [ 14 ]. There are human bones that have been proven reliable for sex estimation and have been examined many times [ 8 , 15 , 16 ]. The most common and reliable of these bones are the most dimorphic skeletal parts such as the pelvis and skull, which form the basis of sex estimation studies [ 17 ]. Of these methods, sex estimation has traditionally been made either by visual assessment based on morphological features of various bones of craniofacial structures or by morphometric methods using linear and/or angular dimensions [ 18 , 19 ]. OA is a section that is particularly noticeable in the head and facial skeleton and has specific differences even among races. Studies on the shape of OA are quite limited and are generally conducted using dry bones [ 2 , 3 , 20 – 23 ]. In our study, we determined the morphology types in coronal slice and 3D OA images and made gender predictions by applying CNN model with deep learning. First, we defined 5 different types of OA and determined the most dominant types according to gender. The most common OA shape in men was round-trapezoid (74%) in both coronal and 3D - MSCT images, while it was round type (39%) in women. After typing, it was recorded that the success rate in coronal images was 65.6% (80% on the side basis) and 73.4% (84.6% on the side basis) in the CNN model in 3D images. Gender identification is essential in forensic anthropology. Research shows that AI techniques like Backpropagation Neural Networks (BPNN) can outperform traditional methods such as Discriminant Function Analysis (DFA) in accuracy, particularly using pelvic and patella data [ 24 ]. BPNN is considered a promising tool for identifying gender, especially in cases involving incomplete or decomposed remains. Studies also indicate that hand bone lengths exhibit sexual dimorphism across populations, though the accuracy of DFA can vary depending on the population [ 14 ] Comprehensive training and the use of 3D models or CT imaging can improve accuracy and reduce observer bias in sex estimation. Estimating height and gender in fragmented remains, such as in mass disasters, presents challenges. To overcome these, new methods are being developed. Zeybek et al. [ 25 ] proposed formulas based on foot measurements for estimating height and gender, while Christina-Schulte et al. [ 26 ] used volumetric CT scans of the skull to estimate age at death and examined sex-based differences. Additionally, studies classifying gender according to the shape of OA are quite limited in the literature [ 27 ] The shape of the OA is determined by the structure of its four edges and the localization of the structures on these edges. This appearance may vary between genders and races. Cameron [ 28 ] stated that the sinus frontalis and sinus maxillaris have an important effect on the formation of the shape of the OA, and that the pulling force of the muscles in the orbital region and the growth of the brain may affect the shape of the OA by affecting the intracranial structure. In the study conducted by Xing et al. [ 2 ] using geometric morphometric analysis method and dry bone in three different populations (European, Asian, African) and determined the orbital shape by examining the superior and inferior margins of OA separately. It was reported that the superior margin was longer in Asians and therefore the orbital shape of Asians was narrower and higher than in Europeans and Africans, whereas it was more inclined and therefore lower than in other populations in Europeans. It was reported that the inferior margin was long and round in Asians, the medial margin was more curved and straight in Europeans, and shorter in Africans than in other populations. In the study of Xing et al. [ 2 ], it was stated that the maxilla and zygomaticum bones forming the inferior margin showed more variability compared to the frontal bone forming the superior margin, and this different growth between the bones caused differences in the shape of OA. Krogman [ 21 ] stated that the shape of OA was angular in Europeans, round in Central Europeans and Asians, and more rectangular in Africans. In addition, it was reported that the OA shape in women was ‘sharp-edged, round, higher and relatively larger’, while in men it was ‘round-edged, square, lower and relatively smaller’. In the study of Komar and Buikstra [ 22 ], it was stated that the OA shape was ‘square, low’ and ‘round-edged’ in men, and ‘round, high’ and ‘sharp-edged’ in women. In the study conducted by Ajanovic [ 20 ] and his colleagues using the geometric morphometric analysis method, the skull was converted into a 3D skull model with a laser scanner. Six reference points were determined, and the obtained shape was compared between the sexes, and it was stated that the orbital shape provided 86.33% accuracy in sex prediction in men and 88.89% in women. Again, in a study conducted using the geometric morphometric analysis method, when the OA shape was evaluated with canonical variance analysis, it was found that the OA shape provided 80% accuracy in women and 73.3% in men [ 29 ]. In the study by Brown and Maeda 23 comparing the skulls of Australian Aboriginals and Tohoku Japanese, it was stated that Australian Aboriginals had dolichocranic cranial vault and rectangular OA, while Tohoku Japanese had brachycranic cranial vault and round OA shape, and it was stated that skull shape may also affect OA shape. In the study conducted by Patra et al. [ 30 ] on digital radiographs of Indian individuals, OA shape was divided into 4 types: round, elliptical, rectangular and square. The most common type was found to be round with 33.5%, followed by elliptical with 30.5%, rectangular with 27.5% and square with 9.5%. When OA shape was compared according to age groups, it was found that there were differences between the sexes in all age groups except the 10–19 age group. It was stated that while the OA shape was round in both boys and girls in the pubertal age group (10–19), it showed sexual dimorphism with age, becoming square and rectangular in boys and elliptical in girls. In our study, the shape of OA was found to be mostly round in the 18–24 age group and round-trapezoidal in the 45–65 age group. Similar to the results of Patra et al. [ 30 ], it was observed that the shape of the OA was round in the early stages of life and its round shape changed in the later stages. It is thought that this may be due to the changes in the maxilla and zygomatic bones forming the margo inferior with age, as stated by Xing et al.[ 2 ] In the study of Lepich et al. [ 4 ], in which they reclassified the OA shape, which Piasecki classified into 15 types, in alphanumeric form, it was concluded that the most common type symmetrically in both males and females was the elongated oval. It was stated that two OA shapes (bottom rounded rhombus-oval round/bottom rounded trapezoid-rectangular oval) were seen asymmetrically in males and asymmetry type was seen in configurations corresponding to elongated oval and oval round types in females. In the study of Çalgüner et al. [ 3 ] on dry bone, OA shape was analysed in 4 types according to Szilvassy classification and it was determined that 53% square, 34% round and 13% trapezoidal shapes were seen in the Turkish population. In our study, it was observed that the OA shape was mostly round-trapezoidal in coronal sectional images, round on the right side and round-trapezoidal on the left side in 3D images. It was also found that the shape differences in the left coronal images were statistically significant. When the age groups were analysed, round-trapezoid was the most common shape in all age groups in coronal section images, while in 3D images, round-trapezoid was the most common shape in the 18–24 age group and round-trapezoid in the 45–65 age group. Similar to the results of Patra et al. [ 30 ], it was observed that the shape of the OA was round in the early stages of life and the round shape changed in the later periods. It is thought that this may be due to the changes in the maxilla and zygomatic bones that form the inferior margin with age, as stated by Xing et al. [ 2 ] Deep learning is an artificial intelligence method that uses multi-layered artificial neural networks and is one of the types of machine learning. Unlike traditional machine learning, they can automatically learn from the symbols of data belonging to images, videos, audio and texts [ 10 ] According to our literature review, although there are studies conducted with machine learning algorithms, it has been observed that the studies on OA morphology using deep learning method are very limited. In the study of Hamwood et al. [ 31 ] automatic segmentation of the OA in CT and magnetic resonance imaging was obtained by deep learning method. They reported that the results were highly compatible with manual segmentation performed by an expert. In addition, it was stated that segmentation of complex structures such as the orbit provides more accurate results in a shorter time by saving time with the deep learning method. In a study by Triantafyllou et al. [ 27 ] using machine learning algorithms on dry bone, a Random Forest Classifier was used and an overall test accuracy of 0.68 was obtained. It was stated that orbital parameters provide reliable results in gender estimation and will contribute to advances in forensic identification techniques. Table 4 shows that studies investigating gender-related variations in structures commonly used in gender determination, such as the skull, OA, sternum, pelvis and foot bones, can be predicted successfully with a success rate of over 90%. Table 4 Comparison of the performance of sex estimation. Study Sample N Method Performance Pretorius [ 29 ] Dried skull orbits Total 60 (50 for each sex) Direct Measurements / Canonical variates analysis 80% for female 73.33% for male Zeybek et al.[ 25 ] Feet Bones Total 498 (249 for each side) Direct Measurements Statistical analysis 95.6% T-test analysis 96.4% Eshak et al.[ 14 ] Hand Bones 2318 Direct Measurements Two and three dimesional reconstruction 92.6% Schule-Geers et al.[ 26 ] Skull 244 Volume, mass, denstity Regression statistics Flat-pamel-based volumetric CT 95.04% Afrianty et al.[ 24 ] Pelvis Patella 136 113 Statistical Measurements Discriminant fuction analysis 86.6% Back propagation neural network 92.9% Darwish et al.[ 32 ] Sternum 4th rib 3D MSCT images 60 Direct Measurements / Multiple regression analysis 96.67% (for sternum) 95.0% (for right 4th rib) 72.68% (for left 4th rib) Uabundit et al.[ 34 ] Dried skull-Pterion landmark 124 Machine learning algorithms / Direct Measurements 80.7% Baban & Mohammed [ 33 ] Mandible CT 3D MSCT 208 (104 for each sex) Machine learning algorithms / Direct Measurements 90% Çiftçi et al.[ 13 ] Skull CT images 421 Deep CNN CNN architectures RelieIF feature selection < 96.4% Triantafyllou et al.[ 27 ] Dried skull orbits 92 Machine learning algorithms / Direct Measurements 68% Current Study Orbits CT and 3D MCT images Total 200 (100 for each side) Deep CNN CNN architectures SVM 65.6% (80% on the side basis, coronal images) 73.4% (84.6% on the side basis, 3D images) In our study, the potential of OA in gender determination was evaluated on coronal and 3D images, and while gender determination was 65.6% (80% on the side basis) in coronal images, it was determined as 73.4% (84.6% on the side basis) in 3D images. This also reveals that 3D images provide more reliable results than coronal images. We believe that our study will be a reference for future studies by increasing the potential of orbital morphology in sex estimation. In CNN modeling, using OA types and each OA together with side information decreased the gender estimation rate. This drew our attention to considering the morphological types of OA and the distribution status between the sides. When the sides are ignored, the gender estimation success can exceed 80% in both coronal and 3D images. Conclusion It is thought that knowing the population-specific shape of the OA, which has significant differences between sexes and races, will provide information about ethnicity in the analyses of skulls from past periods and will be an important source of data in anthropology and forensic medicine in the estimation of identity and gender. In the gender determination analysis performed by deep learning method, it was observed that the accuracy rate of 3D OA images in gender determination was higher. We think that the use of 3D images in gender determination will increase in terms of being close to the real bone appearance, easier and faster access to the desired data, more reliable results can be obtained, and this information will contribute to anthropology and forensic medicine. Declarations Ethics approval and consent to participate The study has been approved by the Board of Medical Research Ethics of Kahramanmaraş Sütçü İmam University (Approval number: 2023/21 & 07). Competing interests The authors declare no competing interests. Funding None. Author Contribution Conceptualization: SAK, MKY, DAS, TK.Methodology: SAK, MKY, DAS, ED.Formal analysis: SAK, MKY, ED, TK.Investigation: SAK, MKY, DAS, GDE, TK.Writing-original draf preparation: SAK, TK.Prepared figures:SAK, GDE, TK.All authors reviewed the manuscript. Data Availability The data that support the findings of this study are available on request from the corresponding author, Karabaş SA, e-mail: [email protected] References Standring, S. (ed) Gray's Anatomy E-Book: Gray's Anatomy E-Book (Elsevier Health Sciences, 2021). Xing, S. et al. Geometric morphometric analyses of orbit shape in Asian, African, and European human populations. Anthropol. Sci. 121 , 1–11 (2013). Çalgüner, E. et al. Anadolu toplumunda adîtus orbitae'nın şekil ve morfometrisi. Turkiye Klinikleri J. Dent. Sci. 3 , 141–145 (1997). Lepich, T. et al. Orbital aperture shape and its alphanumerical classification. Adv. Clin. Exp. Med. 24 , 943–950 (2015). Sarkar, N. & Mukhopadhyay, P. P. Determination of sex from the morphometry of orbits in adult skull of contemporary eastern Indian population. Egypt. J. Forensic Sci. 8 , 61 (2018). Beschiu, L. M. et al. Cranial and odontological methods for sex estimation—A scoping review. Med. (Kaunas) . 58 , 1273 (2022). Spradley, M. K. & Jantz, R. L. Sex estimation in forensic anthropology: skull versus postcranial elements. J. Forensic Sci. 56 , 289–296 (2011). Best, K. C., Garvin, H. M. & Cabo, L. L. An investigation into the relationship between human cranial and pelvic sexual dimorphism. J. Forensic Sci. 63 , 990–1000 (2018). Başer, B. Ö., Yangın, M. & Sarıdaş, E. S. Makine öğrenmesi teknikleriyle diyabet hastalığının sınıflandırılması. Süleyman Demirel Üniv. Fen Bilim Enst Derg . 25 , 112–120 (2021). Yılmaz, A. & Kaya, U. Derin öğrenme. 4. Basım (Kodlab Yayın Dağıtım Yazılım Ltd. Şti., 2022). Lecun, Y. et al. Gradient-based learning applied to document recognition. Proc. IEEE 86, 2278–2324. (1998). https://doi.org/10.1109/5.726791 Cortes, C. & Vapnik, V. Support-vector networks. Mach. Learn. 20 , 273–297 (1995). Çiftçi, R. et al. Human gender estimation from CT images of skull using deep feature selection and feature fusion. Sci. Rep. 14 , 16879 (2024). Eshak, G. A., Ahmed, H. M. & Gawad, E. A. A. Gender determination from hand bones length and volume using multidetector computed tomography: A study in Egyptian people. J. Forensic Leg. Med. 18 , 246–252. https://doi.org/10.1016/j.jfm.2011.04 (2011). Gomez Pellico, L. & Fernandez Camacho, F. J. Biometry of the anterior border of the human hip bone: Normal values and their use in sex determination. J. Anat. 181 , 417–422 (1992). Xue, Y. et al. B. Predicting the risk of acute care readmissions among rehabilitation inpatients: A machine learning approach. J. Biomed. Inform. 86, 143–148. (2018). https://doi.org/10. 1016/j. jbi. 09. 009 (2018). Soler, A. The human skeleton in forensic medicine (Third Edition). By Mehmet Yaşar Işcan and Maryna Steyn. Springfield, IL: Charles C. Thomas. 493 Pp. ISBN 978-0-398-08878-1. $ 70 (Hardcover). Am. J. Phys. Anthropol. 157, 706–707. (2013). https://doi.org/10.1002/ajpa . 22754 (2015). Franklin, D., Cardini, A., Flavel, A. & Kuliukas, A. Estimation of sex from cranial measurements in a western Australian population. Forensic Sci. Int. 229, e1-158.e8. https://doi.org/10. 1016/j. forsc iint. 03. 005 (2013). (2013). Zaafrane, M. et al. Sex determination of a tunisian population by CT scan analysis of the skull. Int. J. Leg. Med. 132 , 853–862. https://doi.org/10.1007/s00414-017-1688-1 (2018). Ajanovic, Z. et al. Geometric morphometrics approach for sex estimation based on the orbital region of human skulls from Bosnian population. Scanning. 2223138 (2023). (2023). Krogman, W. M. The human skeleton in forensic medicine. I. Postgraduate medicine. 17, A-48 (1955). Komar, D. A., Buikstra, J. E. & Forensic Anthropology Contempo-Rary Theory And Practice pp. 362 (Oxford University Press, 2008). Brown, P. & Maeda, T. Post-pleistocene diachronic change in East Asian facial skeletons: the size, shape and volume of the orbits. Anthropol. Sci. 112 (01), 29–40 (2004). Afrianty, I., Nasien, D., Kadir, M. R. & Haron, H. Determination of gender from pelvic bones and patella in forensic anthropology: A comparison of classifcation techniques. In 2013 1st International Conference on Artifcial Intelligence, Modelling and Simulation, 3–7IEEE, (2013). Zeybek, G., Ergur, I. & Demiroglu, Z. Stature and gender estimation using foot measurements. Forensic Sci. Int. 181 (1–3), 54–e1 (2008). Schulte-Geers, C. et al. Age and gender-dependent bone density changes of the human skull disclosed by high-resolution fat-panel computed tomography. Int. J. Leg. Med. 125 , 417–425 (2011). Triantafyllou, G. et al. Sex Estimation Through Orbital Measurements: A Machine Learning Approach for Forensic Science. Diagnostics 14 , 2773 (2024). Cameron, J. Contour of orbital aperture in representatives of modern and fossil Hominidæ. Am. J. Phys. Anthropol. 3 , 476–488 (1920). Pretorius, E., Steyn, M. & Scholtz, Y. Investigation into the usability of geometric morphometric analysis in assessment of sexual dimorphism. Am. J. Phys. Anthropology: Official Publication Am. Association Phys. Anthropologists . 129 , 64–70 (2006). Patra, A. et al. Morphological and morphometric analysis of the orbital aperture and their correlation with age and gender: a retrospective digital radiographic study. Cureus 3 , E17739 (2021). Hamwood, J. et al. A deep learning method for automatic segmentation of the bony orbit in MRI and CT images. Sci. Rep. 11 , 13693 (2021). Darwish, R. T. et al. Sex determination from chest measurements in a sample of Egyptian adults using Multislice Computed Tomography. J. Forensic Leg. Med. 52 , 154–158 (2017). Baban, M. T. A. & Mohammad, D. N. The accuracy of sex identification using CBCT morphometric measurements of the mandible, with different machine-learning algorithms—a retrospective study. Diagnostics 13 , 2342 (2023). Uabundit, N. et al. Classification and morphometric features of pterion in Thai population with potential sex prediction. Medicina 57 (11), 1282 (2021). Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6496345","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":502952954,"identity":"2650d077-5da2-4559-907d-493d35f40bf7","order_by":0,"name":"Sibel Ateşoğlu Karabaş","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABDUlEQVRIiWNgGAWjYDACZh4GhgcQJuMBhooEMEuCoJYEBgMw+wDDGbgWAzx6kLUwthGhRb6d9+CDhJo/cvLt7RcO885Lszc4wHzwNg/Dn3xcWgwO8yUbJBwzMDY4c6bgMO+2nMQNB9iSrXkYDCwbcGlh5jGTSGAzSNwgkZMA1FKRYHCAx0waqAWny+Sbecx/JPwzqJ8/A6RlTgXQYfzf8GphOMxjxpDYZpDAcCP9wGHehhzGDQd42PBqAflFIrHP2HDDmTMMB+ccS0uceZjN2HIO0HM4HdZ/9uCHD9/k5IEh9vDBm5pke77jzQ9vvKmQwxcxMMADVcQMtp0IDQwM7A+IUjYKRsEoGAUjDwAAd9pUavmMfEcAAAAASUVORK5CYII=","orcid":"","institution":"Kahramanmaraş Sütçü İmam University","correspondingAuthor":true,"prefix":"","firstName":"Sibel","middleName":"Ateşoğlu","lastName":"Karabaş","suffix":""},{"id":502952955,"identity":"a3927459-062d-4f4a-a945-5c5e76a39d6d","order_by":1,"name":"Mehlika Küçük Yanar","email":"","orcid":"","institution":"Kahramanmaraş Sütçü İmam University","correspondingAuthor":false,"prefix":"","firstName":"Mehlika","middleName":"Küçük","lastName":"Yanar","suffix":""},{"id":502952958,"identity":"56471e43-556f-4349-9d87-06cdb990bc50","order_by":2,"name":"Duygu Akın Saygın","email":"","orcid":"","institution":"Necmettin Erbakan University","correspondingAuthor":false,"prefix":"","firstName":"Duygu","middleName":"Akın","lastName":"Saygın","suffix":""},{"id":502952963,"identity":"deaf88e9-752a-47c0-a37b-4a8fd2dcb973","order_by":3,"name":"Turan Koç","email":"","orcid":"","institution":"Kahramanmaraş Sütçü İmam University","correspondingAuthor":false,"prefix":"","firstName":"Turan","middleName":"","lastName":"Koç","suffix":""},{"id":502952967,"identity":"e31f9a8a-999b-465c-bda6-502ab77828d8","order_by":4,"name":"Emrah Dönmez","email":"","orcid":"","institution":"Bandırma Onyedi Eylül University","correspondingAuthor":false,"prefix":"","firstName":"Emrah","middleName":"","lastName":"Dönmez","suffix":""},{"id":502952969,"identity":"c0bf50b7-839d-4606-b726-22fefcb15a29","order_by":5,"name":"Ganime Dilek Emlik","email":"","orcid":"","institution":"Necmettin Erbakan University","correspondingAuthor":false,"prefix":"","firstName":"Ganime","middleName":"Dilek","lastName":"Emlik","suffix":""}],"badges":[],"createdAt":"2025-04-21 13:08:41","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6496345/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6496345/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":90306244,"identity":"3616a2e6-60cc-4a84-b594-d0e37efb4904","added_by":"auto","created_at":"2025-09-01 09:26:47","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":78732,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eA representative convolutional neural networks model.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"Fig1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6496345/v1/310ec46c0b8518d6a09762fe.jpg"},{"id":90306251,"identity":"d971bdbb-b1d5-465f-a608-726be15b11f0","added_by":"auto","created_at":"2025-09-01 09:26:48","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":287059,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDetermination of OA shapes in coronal and 3D images.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(OA: Orbital aperture, MS: Maxillary sinus, Type 1: square, Type 2: trapezoid, Type 3: round, Type 4: oval, Type 5: round-trapezoid).\u003c/p\u003e","description":"","filename":"figure2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6496345/v1/5e4826ea79ffd6177c5bcdb2.jpg"},{"id":90306245,"identity":"484de8db-e095-46c2-b1c4-e2c533f16e35","added_by":"auto","created_at":"2025-09-01 09:26:47","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":159251,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEvaluation of confusion matrix in coronal images \u0026amp; Training graph.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(LM: male left OA, LW: female left OA, RM: male right OA, RW: female right OA).\u003c/p\u003e","description":"","filename":"figure3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6496345/v1/6b56e35f0b5a6f732b03d6ec.jpg"},{"id":90307776,"identity":"bb9fc62d-8322-468e-a615-59ccd51b3a97","added_by":"auto","created_at":"2025-09-01 09:34:47","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":153625,"visible":true,"origin":"","legend":"\u003cp\u003eEvaluation of confusion matrix in 3D images \u0026amp; Training graph.\u003c/p\u003e\n\u003cp\u003e(LM: male left OA, LW: female left OA, RM: male right OA, RW: female right OA).\u003c/p\u003e","description":"","filename":"figure4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6496345/v1/79246d5bfad35d6c4ec18530.jpg"},{"id":100133300,"identity":"8a3b667d-8d3c-42a7-9fe8-27fe9190fcb1","added_by":"auto","created_at":"2026-01-13 10:25:34","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1759366,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6496345/v1/d66af52a-63ca-41d2-8fe1-eea857ab5fd3.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Three-Dimensional Morphometric Evaluation of the Orbital Aperture in Multislice Computed Tomography: Anatomical Classification and Deep Learning-Based Sex Estimation","fulltext":[{"header":"Introduction","content":"\u003cp\u003eThe orbital aperture (OA) refers to the opening of the orbit and is an anatomical landmark in the skull. The OA is formed by the frontal, maxillary and zygomatic bones. It has four edges called supraorbital, infraorbital, medial and lateral margins [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. These thickened margins formed by bony structures surround the OA and protect the eye against impacts. The size and shape of the OA varies between populations and genders. These differences are thought to be mainly due to changes in the slope of the superior and inferior margins. It is known that the medial margin is flat and narrow, and the superior margin is more rounded in Asian populations, whereas the superior and inferior margins are more horizontal and parallel in European populations [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. The OA is very important for anthropologists and maxillofacial surgeons, and it shows differences between populations.\u003c/p\u003e\u003cp\u003eWith the developing technology, there is a need to renew the gross anatomical classifications of OA or to evaluate the reliability of existing classifications [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. A good knowledge of the morphology of the OA, which differs between societies, can be used as an important parameter in determining age, sex and race in forensic medicine [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. Sex estimation is an interdisciplinary field requiring expertise used in many fields, especially in forensic medicine and anthropology [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. According to the skeletal parts, the most dimorphic region after the pelvis is considered to be the skull with a success rate of 90% [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. Today, with the widespread use of computer-aided analysis systems, the focus is on systems that can support experts in the decision-making process. For this, machine learning algorithms are often preferred due to their low cost, more reliable and faster results. Machine learning algorithms also increase the predictive power thanks to their ability to combine and manage large and complex data [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Unlike traditional machine learning, image, video, audio and text symbols are processed. Today, its use in the field of medicine is becoming increasingly widespread due to its high success rates in gender and age prediction [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eIn the light of the above information, we set two aims and objectives in our study. The first was to create OA typing in coronal sections and 3D images. Secondly, we aimed to determine the role of deep learning method in gender determination using coronal sections and 3D OA images.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003eStudy Population\u003c/h2\u003e\u003cp\u003e The study was approved by the decision of Kahramanmaraş S\u0026uuml;t\u0026ccedil;\u0026uuml; İmam University Medical Research Ethics Committee No. 2023/21 \u0026amp; 07. In the study, Multislice Computed Tomography (MSCT) images of patients who applied to Necmettin Erbakan University Faculty of Medicine, Department of Radiology between 2020 and 2023 were retrospectively analysed. A total of 100 TMSCT images of 100 individuals (52 females, 48 males) between the ages of 18\u0026ndash;65 years obtained as a result of exclusion criteria (deformity, pathological lesion, trauma or surgical operation history in the orbital, intraorbital and periorbital region, incompletely visualized artefacts and images with low image quality that do not meet the research requirements) were evaluated. In addition, individuals aged 18\u0026ndash;65 years were categorized into four groups as 18\u0026ndash;24, 25\u0026ndash;34, 35\u0026ndash;44 and 45\u0026ndash;65. All participants were given the necessary information about the purpose and importance of the study and were informed that MSCT images would be used. Written consent was obtained from all participants. The study was also conducted in accordance with the principles stated in the Declaration of Helsinki.\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eStudy Design\u003c/h3\u003e\n\u003cp\u003eTemporal 0.625 mm thick temporal MSCT images of the individuals whose heads were taken in the supine position by placing the head in the neutral position with a 256-slice, double-tube MSCT Somatom Drive (Siemens Healthineers, Germany) device was obtained in the PACS system. The imaging parameters of these images were: kV, 80 and 140; mA, 89 and 178; rotation time, 0.28 s; collimation, 256x0.625; FOV, 220 mm. Images were analysed in coronal section using the PACS measurement module. Temporal TMSCT images were saved in Digital Imaging and Communications in Medicine (DICOM) format and 3D images were obtained using RadiAnt DICOM Viewer, an open-source software platform. The images obtained were created by an investigator with 10 years of radiological experience, considering the exclusion criteria.\u003c/p\u003e\u003cp\u003e\u003cem\u003eMorphological Evaluation of\u003c/em\u003e OA\u003c/p\u003e\u003cp\u003eThe shape evaluation of the OA obtained with coronal section and 3D images was performed by 3 different researchers at different times, and then the results were compared, and the disagreeing opinions were re-examined by 3 researchers and a consensus was reached. In the morphological evaluation of OA, a new typing was created by modifying the Szilvassy classification. OA shape was classified into 5 types as type 1: square, type 2: trapezoid, type 3: round, type 4: oval and type 5: round-trapezoid (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\n\u003ch3\u003eDeep Learning\u003c/h3\u003e\n\u003cp\u003eAttributes are vector quantities with a high level of representation that are used in the interpretation of a data. In order to obtain these values, edges, corners, colors, morphological features, etc. are used in image data. Obtaining features in this way is traditional feature extraction. These attributes are calculated and classified with new generation approaches known as convolutional neural networks (CNN), which consist of layers such as input, convolution, normalization, activation, pooling, fully connected layer and output (classification) layer. After the image data is directly given to the CNN model input layer, there is no need for an external feature extraction method. This task is already performed by the filters in the convolution layers of the CNN model. The features calculated up to the fully connected layer are then classified in the classifier layer using the SoftMAX function. CNN models using these features are widely used due to their high performance in the classification task [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eEfficientNet CNN model is one of the new generation convolutional neural network architectures that provides cost advantage as it can control the computational cost more efficiently while increasing the depth and width of the network. The depth, width and resolution values of the network are scaled uniformly, and the instability and inefficiency caused by random scaling are eliminated. This is achieved by using the compound coefficient technique. In this study, the second generation EfficientNet model, EfficientNetv2b0 architecture was used. The number of parameters, which was 5.3M in EfficientNet b0 version, increased up to 7.2M in v2 b0 version. The features were obtained from the fully connected layer of the model. These features were classified with the Support Vector Machine (SVM) classifier using linear and polynomial (quadratic and cubic) Kernel function 44. SVM classifier is a widely used and successful classifier method [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. The method searches for a hyperplane that can best separate the data. SVM classifer can also work if the number of classes is more than two. SVM classifer using linear, quadratic, and cubic kernel functions was used in the study [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. The algorithm attempts to minimise a scalar objective function, f(x), for values of \u0026lsquo;x\u0026rsquo; limited to a finite space. This objective function can be deterministic or stochastic, i.e. it can produce different results when evaluated at the same point x [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. In this study, Expected Improvement (EI) was used to model the criteria for selecting the next hyperparameter. Eq.\u0026nbsp;2 models this function, where, denotes the threshold value of the objective function and \u0026lsquo;\u0026rsquo;y\u0026lsquo;\u0026rsquo; represents the current value of the objective function. The objective is to maximise the value of EI using this equation, taking into account the set of hyperparameters. The equation for the dual version of the general SVM method is given in Eq.\u0026nbsp;(1) below.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eIn the study, data sets were obtained by recording coronal slice and 3D images for deep learning (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Images of the OA were given to the input layer of the EfficientNetv2b0 model. Image features were then extracted from the fully connected layer (dense) of the pre-trained CNN model with 1000 features for each image data. Thus, 987x1000 features were obtained for a total of 987 data, 496 data in each class. These features were trained separately using three different Kernel functions of the SVM classifier. Bayesian optimisation was then used to determine the ideal values of the classifier hyperparameters. Classifier training was completed according to the optimized hyperparameters, and validation was performed by cross-validation. A 10-fold cross-validation approach was used to measure the performance of the trained classifier.\u003c/p\u003e\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\u003ch2\u003eStatistical Analysis\u003c/h2\u003e\u003cp\u003eThe minimum sample size was determined as 70 (α [significance level]\u0026thinsp;=\u0026thinsp;0.05; 1-β [power]\u0026thinsp;=\u0026thinsp;0.95) by power analysis (G*power 3.1.9.4) in accordance with the data obtained from the literature. Statistical analyses were performed with IBM Statistical Package for the Social Sciences (SPSS) Statistics 21.0 (Chicago Illinois).\u003c/p\u003e\u003c/div\u003e"},{"header":"Results","content":"\u003cp\u003eOf the 100 individuals aged between 18 and 65 years included in our study, 48 were male and 52 were female. The mean age was 38.34\u0026thinsp;\u0026plusmn;\u0026thinsp;13 years (male, 36.25\u0026thinsp;\u0026plusmn;\u0026thinsp;11.56; female, 40.27\u0026thinsp;\u0026plusmn;\u0026thinsp;14.04) and the data of the individuals were evaluated in 4 groups as 18\u0026ndash;24 years (20 individuals, mean age 20.80\u0026thinsp;\u0026plusmn;\u0026thinsp;1.85), 25\u0026ndash;34 years (22 individuals, mean age 29.91\u0026thinsp;\u0026plusmn;\u0026thinsp;2.61), 35\u0026ndash;44 years (24 individuals, mean age 40.08\u0026thinsp;\u0026plusmn;\u0026thinsp;2.91) and 45\u0026ndash;65 years (34 individuals, mean age 52.88\u0026thinsp;\u0026plusmn;\u0026thinsp;6.43). The difference between genders in age groups was statistically significant (χ2: 9.213; p\u0026thinsp;=\u0026thinsp;0.027) (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eDistribution by age groups and sex.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"8\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e\u003cp\u003eTotal\u003c/p\u003e\u003cp\u003e(n\u0026thinsp;=\u0026thinsp;100)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003eMale (n\u0026thinsp;=\u0026thinsp;48)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003eFemale (n\u0026thinsp;=\u0026thinsp;52)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAges\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u003cb\u003en\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003e%\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e\u003cb\u003en\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u003cb\u003e%\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e\u003cb\u003en\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e\u003cb\u003e%\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003ep\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e18\u0026ndash;24\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e20\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e20\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e9\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e18.8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e11\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e21.2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e25\u0026ndash;34\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e22\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e22\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e13\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e27.1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e9\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e17.3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e\u003cb\u003e0.027*\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e35\u0026ndash;44\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e24\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e24\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e16\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e33.3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e15.4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e45\u0026ndash;65\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e34\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e34\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e10\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e20.8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e24\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e46.2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e*p\u0026thinsp;\u0026lt;\u0026thinsp;0,05, Chi-square test.\u003c/p\u003e\u003cp\u003eIn observation-based classification, 5 different OA types were determined (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Classification was performed separately for each side. In coronal section images, type 5 (round-trapezoid) was observed most frequently on both sides in both sexes. In 3D images, it was observed that the OA shape was mostly type 5 (round-trapezoidal) in males and type 3 (round) in females, while type 3 (round) was the most common type in the right OA and type 5 (round-trapezoidal) was the most common type in the left OA (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). There was a statistically significant difference between genders in left coronal OA types (p\u0026thinsp;=\u0026thinsp;0.024).\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003e\u003cb\u003eMale and female AO shapes in coronal section and 3D images.\u003c/b\u003e\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"10\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003eTotal (n\u0026thinsp;=\u0026thinsp;100)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003eMale (n\u0026thinsp;=\u0026thinsp;48)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eFemale\u003c/p\u003e\u003cp\u003e(n\u0026thinsp;=\u0026thinsp;52)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003eTypes\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e\u003cb\u003en\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u003cb\u003e%\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e\u003cb\u003en\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e\u003cb\u003e%\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e\u003cb\u003en\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e\u003cb\u003e%\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e\u003cb\u003ep\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"9\" rowspan=\"10\"\u003e\u003cp\u003eCoronal\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"4\" rowspan=\"5\"\u003e\u003cp\u003e\u003cb\u003eRight\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e2.1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e16\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e16\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e10\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e20.8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e6\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e11.5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e11\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e11\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e4.2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e9\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e17.3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e0.149\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e4.2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e1.9\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e69\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e69\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e33\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e68.8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e36\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e69.2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"4\" rowspan=\"5\"\u003e\u003cp\u003e\u003cb\u003eLeft\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e6.3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e12\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e12\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e6\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e12.5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e6\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e11.5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e15.4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e\u003cb\u003e0.024*\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e2.1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e3.8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e74\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e74\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e38\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e79.2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e36\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e69.2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"9\" rowspan=\"10\"\u003e\u003cp\u003e3D\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"4\" rowspan=\"5\"\u003e\u003cp\u003e\u003cb\u003eRight\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e13\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e13\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e7\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e14.6\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e6\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e11.5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e15\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e15\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e16.7\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e7\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e13.5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e35\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e35\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e15\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e31.3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e20\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e38.5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e0.886\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e4.2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e1.9\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e34\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e34\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e16\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e33.3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e18\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e34.6\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"4\" rowspan=\"5\"\u003e\u003cp\u003e\u003cb\u003eLeft\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e14\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e14\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e7\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e14.6\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e7\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e13.5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e17\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e17\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e10\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e20.8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e7\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e13.5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e30\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e30\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e10\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e20.8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e20\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e38.5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e0.411\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e4.2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e3.8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e35\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e35\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e19\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e39.6\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e16\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e30.8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eWhen analysed according to age groups, type 5 was the most common type in coronal measurements in all age groups, while it was found to vary between age groups in 3D images. In the age range of 18\u0026ndash;24 years, type 3 (round) was the most common type of OA on both sides, while type 5 (round-trapezoid) was observed in the age range of 44\u0026ndash;65 years. There was a statistically significant difference between left-sided OA types and age groups (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05) (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003e\u003cb\u003eOrbital types according to age groups in coronal sections and 3D images.\u003c/b\u003e\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"12\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c11\" colnum=\"11\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c12\" colnum=\"12\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colspan=\"8\" nameend=\"c11\" namest=\"c4\"\u003e\u003cp\u003eAge Groups (years)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c12\"\u003e\u0026nbsp;\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003e\u003cb\u003e18\u0026ndash;24\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003e\u003cb\u003e25\u0026ndash;34\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003e\u003cb\u003e35\u0026ndash;44\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c11\" namest=\"c10\"\u003e\u003cp\u003e\u003cb\u003e45\u0026ndash;65\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003eTypes\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e\u003cb\u003en\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u003cb\u003e%\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e\u003cb\u003en\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e\u003cb\u003e%\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e\u003cb\u003en\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e\u003cb\u003e%\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e\u003cb\u003en\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e\u003cb\u003e%\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u003cp\u003e\u003cb\u003ep\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"9\" rowspan=\"10\"\u003e\u003cp\u003eCoronal\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"4\" rowspan=\"5\"\u003e\u003cp\u003e\u003cb\u003eRight\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e4.5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e20\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e6\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e27.3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e6\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e25\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e15\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e4.5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e7\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e20.6\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u003cp\u003e0.052\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e4.5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e2.9\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e12\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e60\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e13\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e59.1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e18\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e75\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e26\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e76.5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"4\" rowspan=\"5\"\u003e\u003cp\u003e\u003cb\u003eLeft\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e4.5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e4.2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e15\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e22.7\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e16.7\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e15\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e14.7\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u003cp\u003e0.136\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e4.5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e2.9\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e12\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e60\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e15\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e68.2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e19\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e79.2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e28\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e82.4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"9\" rowspan=\"10\"\u003e\u003cp\u003e3D\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"4\" rowspan=\"5\"\u003e\u003cp\u003e\u003cb\u003eRight\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e10\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e18.2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e6\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e25\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e2.9\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e9.1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e6\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e25\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e6\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e17.6\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e11\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e55\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e36.4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e20.8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e11\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e32.4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u003cp\u003e0.85\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e8.3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e2.9\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e6\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e30\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e36.4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e20.8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e15\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e44.1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"4\" rowspan=\"5\"\u003e\u003cp\u003e\u003cb\u003eLeft\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e6\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e27.3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e20.8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e5.9\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e20\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e9.1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e6\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e25\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e14.7\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e11\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e55\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e22.7\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e20.8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e9\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e26.5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u003cp\u003e\u003cb\u003e0.011*\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e4.5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e12.5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eType 5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e20\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e36.4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e20.8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e18\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e52.9\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003eSex estimation performance\u003c/h2\u003e\u003cp\u003eCoronal section and 3D images were recorded differently and data sets were created. With deep learning, gender detection in coronal images was 65.6%, while gender detection in 3D images was 73.4%. Deep learning confision matrix and training graphs are given in Fig. s 3 \u0026amp; 4.\u003c/p\u003e\u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eSex estimation is one of the most important milestones in identification [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. There are human bones that have been proven reliable for sex estimation and have been examined many times [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. The most common and reliable of these bones are the most dimorphic skeletal parts such as the pelvis and skull, which form the basis of sex estimation studies [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. Of these methods, sex estimation has traditionally been made either by visual assessment based on morphological features of various bones of craniofacial structures or by morphometric methods using linear and/or angular dimensions [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. OA is a section that is particularly noticeable in the head and facial skeleton and has specific differences even among races. Studies on the shape of OA are quite limited and are generally conducted using dry bones [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan additionalcitationids=\"CR21 CR22\" citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. In our study, we determined the morphology types in coronal slice and 3D OA images and made gender predictions by applying CNN model with deep learning. First, we defined 5 different types of OA and determined the most dominant types according to gender. The most common OA shape in men was round-trapezoid (74%) in both coronal and 3D - MSCT images, while it was round type (39%) in women. After typing, it was recorded that the success rate in coronal images was 65.6% (80% on the side basis) and 73.4% (84.6% on the side basis) in the CNN model in 3D images.\u003c/p\u003e\u003cp\u003eGender identification is essential in forensic anthropology. Research shows that AI techniques like Backpropagation Neural Networks (BPNN) can outperform traditional methods such as Discriminant Function Analysis (DFA) in accuracy, particularly using pelvic and patella data [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. BPNN is considered a promising tool for identifying gender, especially in cases involving incomplete or decomposed remains. Studies also indicate that hand bone lengths exhibit sexual dimorphism across populations, though the accuracy of DFA can vary depending on the population [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e] Comprehensive training and the use of 3D models or CT imaging can improve accuracy and reduce observer bias in sex estimation. Estimating height and gender in fragmented remains, such as in mass disasters, presents challenges. To overcome these, new methods are being developed. Zeybek et al. [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e] proposed formulas based on foot measurements for estimating height and gender, while Christina-Schulte et al. [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e] used volumetric CT scans of the skull to estimate age at death and examined sex-based differences. Additionally, studies classifying gender according to the shape of OA are quite limited in the literature [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]\u003c/p\u003e\u003cp\u003eThe shape of the OA is determined by the structure of its four edges and the localization of the structures on these edges. This appearance may vary between genders and races. Cameron [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e] stated that the sinus frontalis and sinus maxillaris have an important effect on the formation of the shape of the OA, and that the pulling force of the muscles in the orbital region and the growth of the brain may affect the shape of the OA by affecting the intracranial structure. In the study conducted by Xing et al. [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e] using geometric morphometric analysis method and dry bone in three different populations (European, Asian, African) and determined the orbital shape by examining the superior and inferior margins of OA separately. It was reported that the superior margin was longer in Asians and therefore the orbital shape of Asians was narrower and higher than in Europeans and Africans, whereas it was more inclined and therefore lower than in other populations in Europeans. It was reported that the inferior margin was long and round in Asians, the medial margin was more curved and straight in Europeans, and shorter in Africans than in other populations. In the study of Xing et al. [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e], it was stated that the maxilla and zygomaticum bones forming the inferior margin showed more variability compared to the frontal bone forming the superior margin, and this different growth between the bones caused differences in the shape of OA. Krogman [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e] stated that the shape of OA was angular in Europeans, round in Central Europeans and Asians, and more rectangular in Africans. In addition, it was reported that the OA shape in women was \u0026lsquo;sharp-edged, round, higher and relatively larger\u0026rsquo;, while in men it was \u0026lsquo;round-edged, square, lower and relatively smaller\u0026rsquo;. In the study of Komar and Buikstra [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e], it was stated that the OA shape was \u0026lsquo;square, low\u0026rsquo; and \u0026lsquo;round-edged\u0026rsquo; in men, and \u0026lsquo;round, high\u0026rsquo; and \u0026lsquo;sharp-edged\u0026rsquo; in women. In the study conducted by Ajanovic [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e] and his colleagues using the geometric morphometric analysis method, the skull was converted into a 3D skull model with a laser scanner. Six reference points were determined, and the obtained shape was compared between the sexes, and it was stated that the orbital shape provided 86.33% accuracy in sex prediction in men and 88.89% in women. Again, in a study conducted using the geometric morphometric analysis method, when the OA shape was evaluated with canonical variance analysis, it was found that the OA shape provided 80% accuracy in women and 73.3% in men [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. In the study by Brown and Maeda\u003csup\u003e23\u003c/sup\u003e comparing the skulls of Australian Aboriginals and Tohoku Japanese, it was stated that Australian Aboriginals had dolichocranic cranial vault and rectangular OA, while Tohoku Japanese had brachycranic cranial vault and round OA shape, and it was stated that skull shape may also affect OA shape. In the study conducted by Patra et al. [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e] on digital radiographs of Indian individuals, OA shape was divided into 4 types: round, elliptical, rectangular and square. The most common type was found to be round with 33.5%, followed by elliptical with 30.5%, rectangular with 27.5% and square with 9.5%. When OA shape was compared according to age groups, it was found that there were differences between the sexes in all age groups except the 10\u0026ndash;19 age group. It was stated that while the OA shape was round in both boys and girls in the pubertal age group (10\u0026ndash;19), it showed sexual dimorphism with age, becoming square and rectangular in boys and elliptical in girls. In our study, the shape of OA was found to be mostly round in the 18\u0026ndash;24 age group and round-trapezoidal in the 45\u0026ndash;65 age group. Similar to the results of Patra et al. [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e], it was observed that the shape of the OA was round in the early stages of life and its round shape changed in the later stages. It is thought that this may be due to the changes in the maxilla and zygomatic bones forming the margo inferior with age, as stated by Xing et al.[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e] In the study of Lepich et al. [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e], in which they reclassified the OA shape, which Piasecki classified into 15 types, in alphanumeric form, it was concluded that the most common type symmetrically in both males and females was the elongated oval. It was stated that two OA shapes (bottom rounded rhombus-oval round/bottom rounded trapezoid-rectangular oval) were seen asymmetrically in males and asymmetry type was seen in configurations corresponding to elongated oval and oval round types in females. In the study of \u0026Ccedil;alg\u0026uuml;ner et al. [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e] on dry bone, OA shape was analysed in 4 types according to Szilvassy classification and it was determined that 53% square, 34% round and 13% trapezoidal shapes were seen in the Turkish population. In our study, it was observed that the OA shape was mostly round-trapezoidal in coronal sectional images, round on the right side and round-trapezoidal on the left side in 3D images. It was also found that the shape differences in the left coronal images were statistically significant. When the age groups were analysed, round-trapezoid was the most common shape in all age groups in coronal section images, while in 3D images, round-trapezoid was the most common shape in the 18\u0026ndash;24 age group and round-trapezoid in the 45\u0026ndash;65 age group. Similar to the results of Patra et al. [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e], it was observed that the shape of the OA was round in the early stages of life and the round shape changed in the later periods. It is thought that this may be due to the changes in the maxilla and zygomatic bones that form the inferior margin with age, as stated by Xing et al. [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]\u003c/p\u003e\u003cp\u003eDeep learning is an artificial intelligence method that uses multi-layered artificial neural networks and is one of the types of machine learning. Unlike traditional machine learning, they can automatically learn from the symbols of data belonging to images, videos, audio and texts [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e] According to our literature review, although there are studies conducted with machine learning algorithms, it has been observed that the studies on OA morphology using deep learning method are very limited. In the study of Hamwood et al. [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e] automatic segmentation of the OA in CT and magnetic resonance imaging was obtained by deep learning method. They reported that the results were highly compatible with manual segmentation performed by an expert. In addition, it was stated that segmentation of complex structures such as the orbit provides more accurate results in a shorter time by saving time with the deep learning method. In a study by Triantafyllou et al. [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e] using machine learning algorithms on dry bone, a Random Forest Classifier was used and an overall test accuracy of 0.68 was obtained. It was stated that orbital parameters provide reliable results in gender estimation and will contribute to advances in forensic identification techniques. Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e shows that studies investigating gender-related variations in structures commonly used in gender determination, such as the skull, OA, sternum, pelvis and foot bones, can be predicted successfully with a success rate of over 90%.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eComparison of the performance of sex estimation.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"6\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eStudy\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eSample\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eN\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003eMethod\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003ePerformance\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePretorius [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDried skull orbits\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eTotal 60 (50 for each sex)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003eDirect Measurements / Canonical variates analysis\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e80% for female\u003c/p\u003e\u003cp\u003e73.33% for male\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eZeybek et al.[\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eFeet Bones\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eTotal 498 (249 for each side)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eDirect Measurements\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eStatistical analysis\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e95.6%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eT-test analysis\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e96.4%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEshak et al.[\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eHand Bones\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2318\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eDirect Measurements\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eTwo and three dimesional reconstruction\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e92.6%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSchule-Geers et al.[\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eSkull\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e244\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eVolume, mass, denstity\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eRegression statistics\u003c/p\u003e\u003cp\u003eFlat-pamel-based volumetric CT\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e95.04%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eAfrianty et al.[\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003ePelvis\u003c/p\u003e\u003cp\u003ePatella\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003e136\u003c/p\u003e\u003cp\u003e113\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eStatistical Measurements\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eDiscriminant fuction analysis\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e86.6%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eBack propagation neural network\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e92.9%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDarwish et al.[\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eSternum\u003c/p\u003e\u003cp\u003e4th rib 3D MSCT images\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e60\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eDirect Measurements / Multiple regression analysis\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e96.67% (for sternum)\u003c/p\u003e\u003cp\u003e95.0% (for right 4th rib)\u003c/p\u003e\u003cp\u003e72.68% (for left 4th rib)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eUabundit et al.[\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDried skull-Pterion landmark\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e124\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003eMachine learning algorithms / Direct Measurements\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e80.7%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBaban \u0026amp; Mohammed [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMandible CT 3D MSCT\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e208 (104 for each sex)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003eMachine learning algorithms / Direct Measurements\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e90%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u0026Ccedil;ift\u0026ccedil;i et al.[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eSkull CT images\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e421\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eDeep CNN\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eCNN architectures RelieIF feature selection\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;96.4%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTriantafyllou et al.[\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDried skull orbits\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e92\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003eMachine learning algorithms / Direct Measurements\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e68%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCurrent Study\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eOrbits CT and 3D MCT images\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eTotal 200 (100 for each side)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eDeep CNN\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eCNN architectures SVM\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e65.6% (80% on the side basis, coronal images)\u003c/p\u003e\u003cp\u003e73.4% (84.6% on the side basis, 3D images)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eIn our study, the potential of OA in gender determination was evaluated on coronal and 3D images, and while gender determination was 65.6% (80% on the side basis) in coronal images, it was determined as 73.4% (84.6% on the side basis) in 3D images. This also reveals that 3D images provide more reliable results than coronal images. We believe that our study will be a reference for future studies by increasing the potential of orbital morphology in sex estimation. In CNN modeling, using OA types and each OA together with side information decreased the gender estimation rate. This drew our attention to considering the morphological types of OA and the distribution status between the sides. When the sides are ignored, the gender estimation success can exceed 80% in both coronal and 3D images.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eIt is thought that knowing the population-specific shape of the OA, which has significant differences between sexes and races, will provide information about ethnicity in the analyses of skulls from past periods and will be an important source of data in anthropology and forensic medicine in the estimation of identity and gender. In the gender determination analysis performed by deep learning method, it was observed that the accuracy rate of 3D OA images in gender determination was higher. We think that the use of 3D images in gender determination will increase in terms of being close to the real bone appearance, easier and faster access to the desired data, more reliable results can be obtained, and this information will contribute to anthropology and forensic medicine.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003cp\u003eThe study has been approved by the Board of Medical Research Ethics of Kahramanmaraş S\u0026uuml;t\u0026ccedil;\u0026uuml; İmam University (Approval number: 2023/21 \u0026amp; 07).\u003c/p\u003e\u003c/p\u003e\u003cp\u003e\u003ch2\u003eCompeting interests\u003c/h2\u003e\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e\u003cp\u003eNone.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eConceptualization: SAK, MKY, DAS, TK.Methodology: SAK, MKY, DAS, ED.Formal analysis: SAK, MKY, ED, TK.Investigation: SAK, MKY, DAS, GDE, TK.Writing-original draf preparation: SAK, TK.Prepared figures:SAK, GDE, TK.All authors reviewed the manuscript.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThe data that support the findings of this study are available on request from the corresponding author, Karabaş SA, e-mail: [email protected]\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eStandring, S. (ed) \u003cem\u003eGray's Anatomy E-Book: Gray's Anatomy E-Book\u003c/em\u003e (Elsevier Health Sciences, 2021).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eXing, S. et al. Geometric morphometric analyses of orbit shape in Asian, African, and European human populations. \u003cem\u003eAnthropol. Sci.\u003c/em\u003e \u003cb\u003e121\u003c/b\u003e, 1\u0026ndash;11 (2013).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003e\u0026Ccedil;alg\u0026uuml;ner, E. et al. Anadolu toplumunda ad\u0026icirc;tus orbitae'nın şekil ve morfometrisi. \u003cem\u003eTurkiye Klinikleri J. Dent. Sci.\u003c/em\u003e \u003cb\u003e3\u003c/b\u003e, 141\u0026ndash;145 (1997).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLepich, T. et al. Orbital aperture shape and its alphanumerical classification. \u003cem\u003eAdv. Clin. Exp. Med.\u003c/em\u003e \u003cb\u003e24\u003c/b\u003e, 943\u0026ndash;950 (2015).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSarkar, N. \u0026amp; Mukhopadhyay, P. P. Determination of sex from the morphometry of orbits in adult skull of contemporary eastern Indian population. \u003cem\u003eEgypt. J. Forensic Sci.\u003c/em\u003e \u003cb\u003e8\u003c/b\u003e, 61 (2018).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBeschiu, L. M. et al. Cranial and odontological methods for sex estimation\u0026mdash;A scoping review. \u003cem\u003eMed. (Kaunas)\u003c/em\u003e. \u003cb\u003e58\u003c/b\u003e, 1273 (2022).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSpradley, M. K. \u0026amp; Jantz, R. L. Sex estimation in forensic anthropology: skull versus postcranial elements. \u003cem\u003eJ. Forensic Sci.\u003c/em\u003e \u003cb\u003e56\u003c/b\u003e, 289\u0026ndash;296 (2011).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBest, K. C., Garvin, H. M. \u0026amp; Cabo, L. L. An investigation into the relationship between human cranial and pelvic sexual dimorphism. \u003cem\u003eJ. Forensic Sci.\u003c/em\u003e \u003cb\u003e63\u003c/b\u003e, 990\u0026ndash;1000 (2018).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBaşer, B. \u0026Ouml;., Yangın, M. \u0026amp; Sarıdaş, E. S. Makine \u0026ouml;ğrenmesi teknikleriyle diyabet hastalığının sınıflandırılması. S\u0026uuml;leyman Demirel \u0026Uuml;niv. \u003cem\u003eFen Bilim Enst Derg\u003c/em\u003e. \u003cb\u003e25\u003c/b\u003e, 112\u0026ndash;120 (2021).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYılmaz, A. \u0026amp; Kaya, U. \u003cem\u003eDerin \u0026ouml;ğrenme. 4. Basım\u003c/em\u003e (Kodlab Yayın Dağıtım Yazılım Ltd. Şti., 2022).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLecun, Y. et al. Gradient-based learning applied to document recognition. Proc. IEEE 86, 2278\u0026ndash;2324. (1998). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1109/5.726791\u003c/span\u003e\u003cspan address=\"10.1109/5.726791\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCortes, C. \u0026amp; Vapnik, V. Support-vector networks. \u003cem\u003eMach. Learn.\u003c/em\u003e \u003cb\u003e20\u003c/b\u003e, 273\u0026ndash;297 (1995).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003e\u0026Ccedil;ift\u0026ccedil;i, R. et al. Human gender estimation from CT images of skull using deep feature selection and feature fusion. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e, 16879 (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eEshak, G. A., Ahmed, H. M. \u0026amp; Gawad, E. A. A. Gender determination from hand bones length and volume using multidetector computed tomography: A study in Egyptian people. \u003cem\u003eJ. Forensic Leg. Med.\u003c/em\u003e \u003cb\u003e18\u003c/b\u003e, 246\u0026ndash;252. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jfm.2011.04\u003c/span\u003e\u003cspan address=\"10.1016/j.jfm.2011.04\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2011).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGomez Pellico, L. \u0026amp; Fernandez Camacho, F. J. Biometry of the anterior border of the human hip bone: Normal values and their use in sex determination. \u003cem\u003eJ. Anat.\u003c/em\u003e \u003cb\u003e181\u003c/b\u003e, 417\u0026ndash;422 (1992).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eXue, Y. et al. B. Predicting the risk of acute care readmissions among rehabilitation inpatients: A machine learning approach. J. Biomed. Inform. 86, 143\u0026ndash;148. (2018). https://doi.org/10. 1016/j. jbi. 09. 009 (2018).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSoler, A. The human skeleton in forensic medicine (Third Edition). By Mehmet Yaşar Işcan and Maryna Steyn. Springfield, IL: Charles C. Thomas. 493 Pp. ISBN 978-0-398-08878-1. \u003cspan\u003e$\u003c/span\u003e70 (Hardcover). Am. J. Phys. Anthropol. 157, 706\u0026ndash;707. (2013). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/ajpa\u003c/span\u003e\u003cspan address=\"10.1002/ajpa\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. 22754 (2015).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFranklin, D., Cardini, A., Flavel, A. \u0026amp; Kuliukas, A. Estimation of sex from cranial measurements in a western Australian population. Forensic Sci. Int. 229, e1-158.e8. https://doi.org/10. 1016/j. forsc iint. 03. 005 (2013). (2013).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZaafrane, M. et al. Sex determination of a tunisian population by CT scan analysis of the skull. \u003cem\u003eInt. J. Leg. Med.\u003c/em\u003e \u003cb\u003e132\u003c/b\u003e, 853\u0026ndash;862. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s00414-017-1688-1\u003c/span\u003e\u003cspan address=\"10.1007/s00414-017-1688-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2018).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAjanovic, Z. et al. Geometric morphometrics approach for sex estimation based on the orbital region of human skulls from Bosnian population. Scanning. 2223138 (2023). (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKrogman, W. M. The human skeleton in forensic medicine. I. Postgraduate medicine. 17, A-48 (1955).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKomar, D. A., Buikstra, J. E. \u0026amp; Forensic Anthropology \u003cem\u003eContempo-Rary Theory And Practice\u003c/em\u003epp. 362 (Oxford University Press, 2008).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBrown, P. \u0026amp; Maeda, T. Post-pleistocene diachronic change in East Asian facial skeletons: the size, shape and volume of the orbits. \u003cem\u003eAnthropol. Sci.\u003c/em\u003e \u003cb\u003e112\u003c/b\u003e (01), 29\u0026ndash;40 (2004).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAfrianty, I., Nasien, D., Kadir, M. R. \u0026amp; Haron, H. Determination of gender from pelvic bones and patella in forensic anthropology: A comparison of classifcation techniques. In 2013 1st International Conference on Artifcial Intelligence, Modelling and Simulation, 3\u0026ndash;7IEEE, (2013).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZeybek, G., Ergur, I. \u0026amp; Demiroglu, Z. Stature and gender estimation using foot measurements. \u003cem\u003eForensic Sci. Int.\u003c/em\u003e \u003cb\u003e181\u003c/b\u003e (1\u0026ndash;3), 54\u0026ndash;e1 (2008).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSchulte-Geers, C. et al. Age and gender-dependent bone density changes of the human skull disclosed by high-resolution fat-panel computed tomography. \u003cem\u003eInt. J. Leg. Med.\u003c/em\u003e \u003cb\u003e125\u003c/b\u003e, 417\u0026ndash;425 (2011).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTriantafyllou, G. et al. Sex Estimation Through Orbital Measurements: A Machine Learning Approach for Forensic Science. \u003cem\u003eDiagnostics\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e, 2773 (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCameron, J. Contour of orbital aperture in representatives of modern and fossil Hominid\u0026aelig;. \u003cem\u003eAm. J. Phys. Anthropol.\u003c/em\u003e \u003cb\u003e3\u003c/b\u003e, 476\u0026ndash;488 (1920).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePretorius, E., Steyn, M. \u0026amp; Scholtz, Y. Investigation into the usability of geometric morphometric analysis in assessment of sexual dimorphism. \u003cem\u003eAm. J. Phys. Anthropology: Official Publication Am. Association Phys. Anthropologists\u003c/em\u003e. \u003cb\u003e129\u003c/b\u003e, 64\u0026ndash;70 (2006).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePatra, A. et al. Morphological and morphometric analysis of the orbital aperture and their correlation with age and gender: a retrospective digital radiographic study. \u003cem\u003eCureus\u003c/em\u003e \u003cb\u003e3\u003c/b\u003e, E17739 (2021).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHamwood, J. et al. A deep learning method for automatic segmentation of the bony orbit in MRI and CT images. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e11\u003c/b\u003e, 13693 (2021).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDarwish, R. T. et al. Sex determination from chest measurements in a sample of Egyptian adults using Multislice Computed Tomography. \u003cem\u003eJ. Forensic Leg. Med.\u003c/em\u003e \u003cb\u003e52\u003c/b\u003e, 154\u0026ndash;158 (2017).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBaban, M. T. A. \u0026amp; Mohammad, D. N. The accuracy of sex identification using CBCT morphometric measurements of the mandible, with different machine-learning algorithms\u0026mdash;a retrospective study. \u003cem\u003eDiagnostics\u003c/em\u003e \u003cb\u003e13\u003c/b\u003e, 2342 (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eUabundit, N. et al. Classification and morphometric features of pterion in Thai population with potential sex prediction. \u003cem\u003eMedicina\u003c/em\u003e \u003cb\u003e57\u003c/b\u003e (11), 1282 (2021).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Orbit, Multislice Computed Tomography, Deep Learning, Sex estimation","lastPublishedDoi":"10.21203/rs.3.rs-6496345/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6496345/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe orbital aperture (OA) is an important anatomical structure that forms the entrance of the orbit and has connections with intracranial structures. This structure has a critical importance for clinicians as it contains reference points in surgical approaches and requires attention in plastic reconstructive surgery. The OA is also one of the craniofacial variables used for sex estimation in anthropology and forensic medicine. The aim of this study was to evaluate the OA morphologically and to determine its role in sex estimation by deep learning method.\u003c/p\u003e\u003cp\u003eThree-dimensional (3D) images of 100 adult individuals (48 males, 52 females) on Multislice Computed Tomography (MSCT) were used in the study. A classification for the OA shape was created by classifying the right and left side OA images based on observation by 3 researchers. In this study, sex estimation was performed using convolutional neural network (CNN) models to extract deep features found in OA coronal and MSCT images. Gender prediction percentage was calculated using deep learning method over the OA images registered to the algorithm one by one.\u003c/p\u003e\u003cp\u003eFor OA morphology, 5 types were identified: square, trapezoid, round, oval, round-trapezoid. The percentage of gender prediction, with the deep learning method, in coronal slices and 3D images was found to be 65.6% and 73.4%, respectively. The most common OA shape was round-trapezoid (74%) in both coronal and 3D images in males, while it was round type (39%) in females.\u003c/p\u003e\u003cp\u003eIn CNN modeling, incorporating OA types along with side information led to a decrease in gender estimation accuracy. This highlighted the importance of considering the morphological variations of OA and their distribution across sides. Interestingly, when side information is excluded, gender prediction accuracy can exceed 80% in both coronal and 3D images.\u003c/p\u003e","manuscriptTitle":"Three-Dimensional Morphometric Evaluation of the Orbital Aperture in Multislice Computed Tomography: Anatomical Classification and Deep Learning-Based Sex Estimation","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-09-01 09:26:43","doi":"10.21203/rs.3.rs-6496345/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"7bd40e20-e998-483d-81b0-a253e26e0aa3","owner":[],"postedDate":"September 1st, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":53427522,"name":"Health sciences/Anatomy"},{"id":53427523,"name":"Physical sciences/Mathematics and computing"}],"tags":[],"updatedAt":"2026-01-13T10:25:08+00:00","versionOfRecord":[],"versionCreatedAt":"2025-09-01 09:26:43","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6496345","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6496345","identity":"rs-6496345","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}