Mandible: An Indicator for Sex Determination - A Three-dimensional Cone-Beam Computed Tomography Study.

BACKGROUND: Mandible is considered as one of the stronger bones of skull available for gender identification. Mandibular measurements can be used for the identification of gender either on dry mandible or through panoramic radiography or cone-beam computed tomography (CBCT). AIM: To determine the gender from mandibular measurement using CBCT.
MATERIALS AND METHODS: The Morphometric analysis was performed on 200 CBCT scans of the subjects using OnDemand 3D software (Seoul, Korea). Morphometry of mandibular is measured by estimation of the angles formed at different locations on mandible to predict the gender. Statistical analysis was performed; independent samples "t-" test was used to compare the mean values between males and females. Discriminant function analysis was used for gender prediction.
RESULTS: The subject's age ranged from 18 to 60 years with an equal number of males and females. The mean angle formed by the intersection of lines from the left and right gonion to menton in males was 129.9 ± 11.9 and it was 126.7 ± 12.6 in females. The mean linear distances from the right gonion to menton in both males and females were 86.8 ± 5.3 and 82.6 ± 6.2, respectively. The mean linear distances from the left gonion to menton were 49.5 ± 5.1 in males and 47.7 ± 3.9 in females. The mean linear distance from the right gonion to left gonion was 47.7 ± 4.4 in males and 46.6 ± 4.2 in females. The Box's M statistics was applied to verify the applicability of mandibular measurements in gender prediction. The values indicate that gender can be predicted using these four variables, which is statistically significant (P = 0.000). The overall prediction accuracy of this model was 67%, with 66.7% in males and 67.3% in females being correctly classified.
CONCLUSION: The angle formed by the intersection of lines from the left and right gonion to menton (Gn-M0) helps provide anthropological data, which can be used in dental and medicolegal practices. Copyright:

Introduction

Forensic dentistry gained its importance in victim identification in criminal investigations, mass disasters, and war crimes.[1] The precision in identification of human skeletal is vital for forensic identification of victim. Gender identification from the skeletal remnants was considered as one of the foremost steps in forensic identification.[2] Pelvic bones were considered as the most accurate bones for gender identification; however, pelvic bones are one of the weakest bones and may not be available in many cases.[3] Mandible is considered as one of the stronger bones of skull available for gender identification. Gender identification is mainly based on the morphological characteristics of mandible, and radiology plays an important role in it.[3]

Using conventional two-dimensional (2D) cephalometric analysis may lead to bias due to difficulties in finding accurate measurement point due to overlapping of bony structures.[4] To overcome this problem, 3D modality such as 3D cone-beam computed tomography (CBCT) has been suggested. CBCT has been revealed to be a superior tool than conventional imaging modalities for the depiction of the bone morphology; however, the visibility of this structure may vary significantly, even within the same individual.[567] Discriminant function analysis is the commonly used statistical procedure in gender identification. It is population specific, and different populations need different regional criteria or formulae in using discriminant function analysis.[8] As literature search could not identify studies using 3D modality in gender identification among population of Saudi Arabia, the study aimed to assess the usefulness of 3D radiographic analysis of mandible in gender identification among population of Saudi Arabia.

Materials and Methods

Ethical clearance had been obtained from the local committee of bioethics with the approval number of 9-16-8/39.

Study design

This retrospective study was conducted over a period of 6 months from the CBCT scans obtained from the Radiology department, College of Dentistry, Al Jouf University. CBCT scans were obtained from the OnDemand 3D software (Seoul, Korea). The samples were selected based on the systematic random sampling technique. The inclusion criteria included the age group of 18–60 years, full complement of teeth in mandible, and systemically healthy. Patients with a history of bone disorders, severe developmental anomalies, and the presence of any bony lesions leading to variation in morphology of mandible were excluded from the study. Morphometry of mandibular is measured by estimation of the angles formed at different locations on mandible to predict the gender as shown in Figure 1.

Figure 1

Mandibular morphometry for the gender determination. (Upper row for female cone-beam computed tomography image and lower row for male cone-beam computed tomography image). A: Angle formed by gonion right to menton to gonion left. B: Linear distance from the gonion right to menton. C: Linear distance from the gonion left to menton. D: Linear distance from the gonion right to gonion left

Sample size calculations

The sample size was based on the recommendations of Hair et al.,[9] who stated that at least fifty samples are needed for each independent variable in multiple discriminant analysis. In this study, four independent variables were considered in discriminate analysis (4 × 50 = 200), so the final sample size was determined as 200.

Statistical analysis

Collected data were analyzed using Statistical Package for the Social Sciences (SPSS, version 20, Chicago, Inc., IL, USA). Descriptive statistics were done initially; independent samples “t-” test was used to compare the mean values between males and females. Discriminant function analysis was used for gender prediction and P ≤ 0.05 was considered statistically significant for all comparisons.

Results

The age of the study participants ranged from 18 to 60 years with the mean age of 39 (±2.05) years. Of 200 participants, 96 were male and 104 were female. Table 1 describes descriptive statistics of various variables in both genders. The mean angle formed by the intersection of lines from the left and right gonion to menton in males was 129.9 ± 11.9 and it was 126.7 ± 12.6 in females. The mean linear distances from the right gonion to menton in both males and females were 86.8 ± 5.3 and 82.6 ± 6.2, respectively. The mean linear distances from the left gonion to menton were 49.5 ± 5.1 in males and 47.7 ± 3.9 in females. The mean linear distance from the right gonion to left gonion was 47.7 ± 4.4 in males and 46.6 ± 4.2 in females. The Box's M statistics was applied to verify the applicability of mandibular measurements in gender prediction. The values indicate that the gender can be predicted using these four variables, which is statistically significant (P = 0.000) [Table 2]. Unstandardized coefficient values obtained from the Canonical Discriminant Function Coefficients [Table 2] were used in gender prediction. The estimated sex was calculated using the following equation: N = −17.3+ (0.31 × Gonion-Menton) + (0.149 × right Gn-M) + (0.86 × left Gn-M) + (−0.07 × right Gn-left Gn). The sectioning (Eigen) value for gender prediction was 0.41; if the calculated value is 0.41 or above, then that mandible belongs to male and for females it is below 0.41 [Table 3]. Considering the four independent variables, 66.7% of males were correctly classified and 67.3% of males were correctly classified, and the overall prediction accuracy of this model is 67% [Table 4].

Table 1

Discriminant analysis of various parameters of mandible

Particulars	Parameters	Mean±SD	P
Male	The angle formed by the intersection of lines from the left and right gonion to menton	129.9±11.9	0.02
	Linear distance from the right gonion to menton	86.8±5.3
	Linear distance from the left gonion to menton	49.5±5.1
	Linear distance from the right gonion to left gonion	47.7±4.4
Female	The angle formed by the intersection of lines from the left and right gonion to menton	126.7±12.6	0.000
	Linear distance from the right gonion to menton	82.6±6.2
	Linear distance from the left gonion to menton	47.7±3.9
	Linear distance from the right gonion to left gonion	46.6±4.2
Overall	The angle formed by the intersection of lines from the left and right gonion to menton	128.2±12.3	0.005
	Linear distance from the right gonion to menton	84.6±6.1
	Linear distance from the left gonion to menton	48.6±4.6
	Linear distance from the right gonion to left gonion	47.2±4.3

SD: Standard deviation

Table 2

Box’s M statistics for gender identification

Box’s M	F
	Approximate	df1	df2	P
35.42	3.46	10	18.46	0.000

Table 3

Canonical discriminant function

Parameters	Unstandardized coefficients	Sectioning point
		Male	Female
The angle formed by the intersection of lines from the left and right gonion to menton (Gn-M0)	0.031	0.418	−0.418
Linear distance from the right gonion to menton (Rt Gn-M)	0.149
Linear distance from the left gonion to menton (Lt Gn-M)	0.086
Linear distance from the right gonion to left gonion (Rt Gn-Lt Gn)	−0.072
Value of constant	−17.300

Table 4

Prediction accuracy

Particulars	Prediction accuracy (%)
Male	66.7
Female	67.3
Overall	67

Discussion

Sex determination of adult skeleton is usually the first step of the identification process, as subsequent methods for age and stature estimation are sex dependent. Jaws and teeth have been used since olden times to ascertain the sex of an individual. The mandible is the largest and toughest bone in the face with a horizontally curved body that is convex forward with two broad rami, which ascend from the posterior end of the body. The mandible is considered suitable for the study as it is the most durable bone of the facial skeleton and retains its shape better than other bones. Sexual dimorphism in the mandible may be due to the relative difference in the development of the musculoskeletal system, especially the muscles of mastication attached to the mandible.[10]

In the present study, the angle formed by the intersection of lines from the left and right gonion to menton was evaluated to determine the sex dimorphism which ranged from 129.9 ± 11.9 in males and 126.7 ± 12.6 in females. These findings were in accordance with the study conducted earlier on the angle of the mandible on a mixed population by various researches.[1112] The angle varied between 110° and 140°. It was also concluded that in the patients who retained their teeth, there is no tendency of increase in the angle with advancing age. The findings from the present study revealed that using the measured parameters on mandible, a prediction rate of 66.7% for males, 67.3% for females, and overall 67% could be projected, which is very low in comparison with the prediction rate proposed by Abu-Taleb and El Beshlawy.[13] The difference in such variation could be with the aspect of methodology and more accurate measurements were done on CBCT. Various researchers conducted studies on dry adult mandibles of known sex using anthropometric measurements and found an accuracy rate ranging from 69% to 94% and reported that ramus height and breadth were highly significant parameters. A detailed comparison of the findings from the present study can be analyzed with the various studies done by researches listed in Table 5. Major limitations of the study are sample size, arbitrary measurement, and analysis of the mandible, which are subject to the whim of each examiner and dental profile changes over a period of time. More research is needed to confirm these findings on cadavers.

Table 5

Comparison of various studies for gender prediction accuracy

Authors and year	Nationality	Parameters studied	Mode of study	Observations
More et al.[1]	Indian	Mandibular ramus	Digital OPG	The overall accuracy for diagnosing gender was 69%, whereas for diagnosing male and female, the accuracy was 68% and 70%, respectively
Maloth et al.[14]	Indian	Upper ramus breadth, lower ramus breadth, condylar ramus height, projective ramus height, and coronoid ramus height	Digital panoramic radiographs	The overall accuracy for diagnosing gender was 74%, whereas for diagnosing male and female, the accuracy was 70% and 78%, respectively
Bangi et al.[15]	Indian	Mediolateral, superoinferior, and anteroposterior dimensions and the volume of the maxillary sinuses	Computed tomography	The accuracy rate was found to be 84% in males and 92% in females, with the mean accuracy of 88%
Khaitan et al.[16]	Indian	Size of the maxillary sinuses	Digital lateral cephalometric radiographs	A sensitivity of 68% and specificity of 76% was observed
Gamba Tde et al.[17]	Brazilian	Ramus length, gonion-gnathion length, minimum ramus breadth, gonial angle, bicondylar breadth, and bigonial breadth	Cone-beam computed tomography	The rate of correct gender classification was 95.1%. An accuracy of 93.33% and 94.74% was found for estimating males and females, respectively
Sairam et al.[18]	Indian	Measurements of mandible	Digital panoramic radiographs	The accuracy of sex determination in mandibular ramus measurements was 79.5% on the right side and 77% on the left side, whereas, in linear mandibular measurements, it was 76% on the right side and 79.5% on the left side
Tambawala et al.[19]	Indian	Foramen magnum dimensions	Cone-beam computed tomography	Overall accuracy rate of sex determination was 66.4%. Of these, 70.3% of males and 62.6% of females were differentiated correctly
Abu-Taleb and El Beshlawy et al.[13]	Egyptian	Mandibular ramus linear and gonial angle measurements	Digital OPG	Accuracy of 81% in males and 77.9% in females and an overall accuracy of 79.6% were noted
Kharoshah et al.[20]	Egyptian	Osteometric mandibular measurements	UAO and DPT radiographs	The overall predictive accuracy was 83.9%. The correct predictive accuracy was 83.6% in males and 84.2% in females
Hsiao et al.[21]	Taiwanese	Cephalometric parameters	Lateral radiographic cephalometric	100% accuracy in sex determination

OPG: Orthopantomogram; UAO: Upper anterior occlusal; DPT: Dental panoramic tomographs

Conclusion

The angle formed by the intersection of lines from the left and right gonion to menton (Gn-M0) helps in providing anthropological data, which can be used in dental and medicolegal practices.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.