Journal of Forensic Science and Medicine

: 2022  |  Volume : 8  |  Issue : 2  |  Page : 46--51

A retrospective study to evaluate the morphometry of the foramen magnum and its role in forensic science in a nigerian population of Delta State

Beryl Shitandi Ominde, Patrick Sunday Igbigbi 
 Department of Human Anatomy and Cell Biology, Delta State University, Abraka, Nigeria

Correspondence Address:
Beryl Shitandi Ominde
Department of Human Anatomy and Cell Biology, Delta State University, P.M.B. 1, Abraka


Background: Sexual dimorphism of the foramen magnum has increased its interest in forensic science. Gender determination is an important preliminary step in the identification of unknown skeletal remains. This study aimed at determining the dimensions of the foramen magnum in Delta State Nigeria and their role in gender discrimination. Materials and Methods: We retrospectively analyzed computed tomographic images of 336 patients (199 males and 137 females) aged ≥18 years, archived in the Radiology Department of a Teaching Hospital in Nigeria. Ethical approval was granted by the hospital's ethical board. The length, width, and area of the foramen magnum were measured and analyzed using the Statistical Package for the Social Sciences software version 23. We used an independent t-test and analysis of variance to evaluate the association of these dimensions with sex and age, respectively. The percentage accuracy of sex discrimination and the association between variables were assessed using discriminant functional analysis and Pearson's correlation test correspondingly. The results were considered significant at P < 0.05. Results: The foramen magnum length, width, and area showed a statistically significant gender difference (P < 0.05). The width was the best sex discriminating variable (64.3%) and the overall accuracy of correct sex allocation using all the variables was 75%. All the parameters measured showed a significant strong positive correlation with each other (0.5 ≤ r < 1, P < 0.05). Conclusion: The foramen magnum length width and area were sexually dimorphic. Their high overall accuracy (75%) in gender discrimination implies that they may collectively be utilized in the sex estimation of unknown skulls in Delta State Nigeria.

How to cite this article:
Ominde BS, Igbigbi PS. A retrospective study to evaluate the morphometry of the foramen magnum and its role in forensic science in a nigerian population of Delta State.J Forensic Sci Med 2022;8:46-51

How to cite this URL:
Ominde BS, Igbigbi PS. A retrospective study to evaluate the morphometry of the foramen magnum and its role in forensic science in a nigerian population of Delta State. J Forensic Sci Med [serial online] 2022 [cited 2022 Aug 18 ];8:46-51
Available from:

Full Text


Unknown skeletal remains retrieved following mass disasters such as fires, air crashes, and explosions require a thorough forensic anthropological examination for positive identification.[1] Bone is highly resistant to high temperatures and extreme decomposition, hence reliable in forensic identification.[2] Moreover, it is ideal for constructing the biological profile of unknown individuals such as age, gender, and race estimation that aid in personal identification.[1],[3] The skull bone is thick and withstands physical insults and inhumation hence plays a crucial role in identification.[2]

The occipital bone is a constituent of the skull base that houses the foramen magnum on its basilar part (FM)[1] Evolution brought about the forward migration of the FM further underneath the head in bipeds such as humans.[4] The skull base has abundant soft-tissue coverage which makes it resistant to physical damage hence making the FM ideal for use in forensic science.[1],[5] This foramen transmits the medulla oblongata, meninges, vertebral arteries, anterior and posterior spinal arteries, and spinal accessory nerves.[1] During the embryonic development of the skull, the FM forms by the fusion of four primary cartilaginous centers that form the occipital bone encircling the medulla oblongata. These include the pars squama posteriorly, right and left pars lateralis and the pars basilaris anteriorly.[6] The complete growth of the foramen is achieved in early childhood; hence, it does not respond to the secondary sexual changes and remains stable beyond adolescence.[7] Furthermore, FM shape and size are not altered by the influence of surrounding musculature.[8]

The determination of gender is preliminary in forensic identification of human remains and requires accurate and reliable sex markers.[5] Gender determination narrows the probability of correct identification to 50% since half of the population is eliminated from the pool of possible victim matches.[1],[9] Nonmetric anthropological parameters used to evaluate for sexual dimorphism are subjective and prone to inter-observer errors. Craniometry has subsequently been universally adopted as an objective, systematic, and reproducible method which complements the visually assessed traits.[10] The skull morphometric parameters provide clues that aid in gender estimation with a 90% accuracy.[3],[9]

The use of imaging such as X-ray, computed tomography (CT), and magnetic resonance imaging in skull morphometry provides data consistent with the measurements on dry skull using calipers.[11] Radiographic measurements are more reliable since they are not limited by flesh barrier on the skull.[12] They are also preferred in the analysis of decomposed and charred remains since they eliminate the need for maceration of bones which may make the anthropological examination time-consuming.[2],[7] The use of CT in forensics offers high resolution and three-dimensional display of structures that enhance correct location of the craniometrics points, for accurate morphometry[11] The high demand of CT in routine medical services currently provides a database of images that may be utilized in the derivation of accurate formulas to aid in forensic investigations.[13]

The FM has various medico-legal applications in cases of massive skeletal relics in mass disasters.[11] Its dimensions vary in different population groups owing to the different geographical, ethnic, sex, racial, genetic, and environmental factors.[4],[14],[15],[16] They are useful to anthropologist and morphologists for the estimation of gender, ethnicity, and stature.[4],[15] The sexual dimorphism of the metric parameters of the FM is population-specific.[2],[7] This significant association with gender has been documented in various studies.[1],[2],[11] However, this was not the case in the Kenyan skulls evaluated by Loyal et al.[14]

The discriminant function analysis (DFA) is a statistical model that has been widely adopted by forensic scientists to predict group membership from variables.[2],[8] The use of DFA in the prediction of sex using FM dimensions has shown varying accuracies in different populations.[2],[7],[8],[17] The overall accuracy of gender estimation using the length and width of the FM was 70.9% and 90% in Egypt and Brazil correspondingly.[11],[17] In addition to these parameters, the FM area was included as a sex predictor variable and this yielded a percentage accuracy of 69.7%−80% in India.[5],[7] These discrepancies exist because discriminant formulas are sensitive to the variations in a given population.[10] This study, therefore, aimed at determining the dimensions of the FM in a Nigerian population and further elucidate the accuracy of using these measurements in gender determination.

 Materials and Methods

This retrospective study was conducted in the Radiology Department of a Teaching Hospital in Delta State, Nigeria. Ethical approval was granted by the Hospital's Research and Ethics Committee (EREC/PAN/2020/030/0371). We used brain CT images of 336 adult patients (199 males and 137 females), aged 18 years and above. These images were acquired using a 64 slice CT scanner (Toshiba Aquilon, Japan, 2009) at 120 kV and 300 mA in 5 mm axial slices. We used images taken over a 5-year duration; between June 01, 2015, and June 30, 2020, and stored in the picture archiving communications systems (PACSs). These were images of patients with chronic headache, suspected stroke, space-occupying lesions and pulmonary embolism who were referred for further radiological evaluation. Our exclusion criteria entailed; CT images of patients aged below 18 years, poor quality images such as those with artifacts or patient rotation and images with evidence of congenital anomalies, fractures, and past surgeries of the skull base region.

Using a bone window, we identified the FM on the axial slices. Its length (anteroposterior [AP] diameter) and width were measured using the measuring tool provided by the PACS and calibrated in cm. The FM length was the AP distance (AP diameter) from the basion to opisthion at the midsagittal plane while the width was measured as the maximum transverse diameter between the lateral margins and perpendicular to FM length [Figure 1].[12] The FM area was calculated by PACS after tracing the bony border of the FM on the axial section [Figure 2].{Figure 1}{Figure 2}

To substantiate the reproducibility of the methodology used, we performed the inter-observer and intra-observer concordance tests using forty randomly selected CT images. Two observers independently performed the inter-observer analysis in 20 images at different times. A single observer applied the measurement procedure in a separate set of 20 images to test the intra-observer agreement. This was applied at different times considering a 1-month interval between measurements. After performing the concordance test, the measurements in all the images were performed by a single investigator.

The Statistical Package for the Social Sciences (SPSS) software version 23; IBM® Armonk, New York, USA was used to perform the data analysis. Initially, data were classified according to gender and 10-year age groups [Table 1]. The measurements were summarized in means and standard deviations. An Independent t-test was used to determine the gender differences in the FM dimensions. The association between the quantitative variables and age was probed using the Analysis of variance (ANOVA) test while the correlation between the morphometric parameters was evaluated using Pearson's correlation test. Statistical significance was considered at P < 0.05.{Table 1}

We applied the DFA to determine the accuracy of predicting gender using the FM morphometric parameters. DFA was preferred since it is an objective technique that uses the minimum number of predictors to obtain a highly effective and reliable equation for distinguishing two groups.[11] Its assumptions include multivariable normality, absence of outliers, absence of multicollinearity among predictors, and homogeneity of variance-covariance matrices within groups.[18] This study employed the Kolmogorov–Smirnov test to assess the normality of the distribution. The equality of covariance matrices was assessed using Box's M test.

Coefficients and constants were calculated to derive a discriminate function equation. All the three measured variables of the FM were assessed using multivariate analysis. Furthermore, each variable was subjected to the univariate analysis. The outcome was cross validated using the “leave one out classification” analysis. We determined the discriminant functional scores of male (Dm) and female (Df) FM by applying the mean values of both genders using the equation; Discriminant functional score (D)=(b0)+(b1X1)+(b2X2)+(b3X3) where b0 was a constant, b1–b3 were coefficients and X1-X3 were the metric variables. The average values of the mean of both male and female variables were employed in determining the sectioning point (D0) which was used for gender discrimination. When D value was above D0, the skull was classified male, and when less than the D0, it was considered female. Wilk's lambda in ANOVA (F) tests the differences of means in DFA. A variable with a smaller lambda indicated its high contribution to the discriminant function. Normally, lambda ranges from 0 to 1, with a value closer to zero strongly discriminating between males and females.[10]


The correlation coefficients (r = value) obtained in the inter-observer and intra-observer analysis did not show any statistically significant differences between the measurements (FM AP r = 0.583, P = 0.142; FM width r = 0.567, P = 0.224; FM Area r = 0.511, P = 0.337) (FM AP r = 0.551, P = 0.411; FM width r = 0.542, P = 0.158; FM Area r = 0.530, P = 0.356). This study evaluated the metric parameters of the FM using brain CT images of 336 adult patients with a larger proportion being males (199, 59.2%) compared to 137, 40.8% females the age of the patients ranged from 20 years to 99 years and the mean age was 53.29 ± 18.18 years. The distribution of patients in each age group is shown in [Table 1].

The AP diameter of the FM averagely measured 3.470 ± 0.550 cm; 3.472 ± 0.649 cm in males and 3.368 ± 0.363 cm in females. The maximum transverse width of the FM measured 3.080 ± 0.561 cm; 3.076 ± 0.427 cm and 3.087 ± 0.714 cm in females and males, respectively. The length was significantly larger than the width (P = 0.021). The area of the FM averagely measured 6.470 ± 1.537 cm2 and was larger in males (6.695 ± 1.618 cm2) than in females (6.136 ± 1.350 cm2). The FM length, width, and area showed a significant association with gender (P = 0.039, 0.026, 0.001). However, these variables did not show any significant differences between the age groups (P = 0.399, 0.621, and 0.453) [Table 2]. These three metric parameters also showed a significant strong positive correlation with each other (0.5 ≤ r <1, P < 0.05) [Table 3].{Table 2}{Table 3}

The sample was ideal for developing formulas for estimating gender as evidenced by a normal distribution of variables in the Kolmogorov–Smirnov test (P = 0.084) and equality of covariance matrices demonstrated in the Box's M test (P = 0.001). In the univariate analysis, each FM variable was considered for gender estimation, and the sectioning points used for gender discrimination are shown in [Table 4]. The best variable for gender discrimination was the FM width (216, 64.3%). Using each variable, the original accuracy of prediction was equivalent to the rate after cross-validation [Table 5].{Table 4}{Table 5}

Using multivariate analysis, we derived the following equation; Discriminant functional score (D) = −0.338* −1.344*FM AP diameter − 0.093* FM width + 0.821*FM Area. The canonical coefficients for each parameter were used in the equation to aid in the classification. The sectioning point (D0) was calculated as 0.045. Males were predicted when D > 0.045 while females were predicted when D < 0.045 [Table 6]. The overall accuracy of correct sex allocation was 75% (252 subjects) including 155 (77.9%) males and 97 (70.8%) females. This accuracy reduced after cross validation (240, 71.4%) [Table 5].{Table 6}


The ICC obtained in both the intraobserver and interobserver analysis signifies a moderate reproducibility and repeatability of the measuring protocol of all variables herein.[19] The mean AP diameter of the FM was 3.470 ± 0.550 cm, which was lower than reports from previous CT studies in India and South Africa.[15],[16] Direct measurement of this variable on the skull bone revealed larger FM lengths in Nigerians, Kenyans, and Turkish and smaller AP dimensions in Indians.[4],[14],[20],[21] The discrepancies in the AP diameters in different populations may be ascribed to the differences in race and geographical location. The variation within a population group may be linked to the genetic differences. The differences in the methodology; direct anatomical versus radiographic measurements, sample size, and composition besides the error of measurements, may have contributed to the variations in literature.[12] Congruent with previous literature reports, the FM length was significantly larger than the width.[16],[22]

The width of the FM measured 3.080 ± 0.561 cm and this was higher than the width documented in Indian and South African CT studies.[15],[16] In Nigerian skull bones, larger FM width was documented by Eboh.[4] Similarly, direct skull bone measurements in India and Turkey also revealed smaller FM widths, while in Kenyans, this parameter was larger than our findings.[14],[20],[21] These discrepancies could be attributed to the variations in sample size, methodology, genetics, race, geographical location, environmental, and socioeconomic factors.

The average area of the FM (6.470 ± 1.537 cm2) was lower than the area reported in Nigeria, South Africa, and India.[4],[16],[20] The variation in this parameter may similarly be attributed to the differences in genetics, race, sample size, and methodologies used. The knowledge of the FM area is important since smaller (stenosed) FM is allied with atlas occipitalization and skull base underdevelopment.[23] In achondroplasia, the FM is small essentially due to diminished bone growth caused by abnormal endochondral bone growth and premature fusion of the synchondrosis.[22]

FM length, width, and area herein showed a statistically significant association with gender. This corresponded with Lyrtzis et al.,[23] Gargi et al.[15] and Abo El-Atta et al.[11] who observed significantly larger FM dimensions in males than in females. Conversely, Loyal et al.[14] did not observe any significant gender differences in the metric parameters of the FM. This lack of sexual dimorphism in some populations can be attributed to this foramen reaching its adult size in early childbood. Unlike the innominate bone and the long bones, the FM does not respond to the growth spurt that simultaneously occurs with the significant secondary sexual changes.[7]

Gender determination of the human crania is based on the size differences and robusticity which may be attributed to socioeconomic status, environmental factors, and genetic elements.[7] The FM size is altered by secondary mechanisms including the weight of the head. The males have heavier brains than females and this corresponds to a higher load that increases the mechanical forces transmitted at the atlantooccipital joint.[22] No muscle acts upon the FM; however, it is preserved by the attachment of muscles that are larger in males than females, amounting to wider FM surface area in males.[23]

The best sex discriminating variable in the present study was the FM width (64.3%). This corresponded with the reports of Abo El-Atta et al.[11] in Egypt. Conversely, the AP diameter of the FM was the most accurate tool for sexual differentiation in CT studies by Lopez-Capp et al.[17] and Wani et al.[5] Some studies in the literature showed that the area of the FM had the highest probability for correct sex allocation while our study reports the least accuracy with this variable (57.4%).[6],[7] According to the multivariate analysis using the three metric variables, the probability of accurate gender estimation was 75%. This was higher than the findings of Vinutha et al.,[8] Abo El-Atta et al.,[11] and Verma et al.[7] In contrast, Lopez-Capp et al.[17] and Wani et al.[5] reported higher accuracies compared to our findings. The discrepancies in the percentage accuracy may perhaps be ascribed to the differences in genetics, race, sample size, and statistical tools used such as DFA and Logistic regression analysis in the different studies.[4],[11] The comparison of findings from various literature sources reveals that population differences contribute to the sexual dimorphic features of the cranium hence, the source of an unidentified skull should be acknowledged to adopt a method of sexual determination such as appropriate formulas, based on data from that specific population.

Conforming to the reports of Akay et al.[12] and Abo El-Atta et al.,[11] we report no significant differences in the FM dimensions between the age-groups studied (P > 0.05). According to Lucena et al.,[22] the adult length of the FM is attained by the 5th year of life while its breadth continues to grow until the completion of the first decade. The dimensions of the FM, therefore, remain constant after the second decade of life and its variations are not largely due to age. The medulla oblongata largely occupies the FM and matures at a younger age before the skeletal system hence, the FM does not require to increase in size with age.[7]

Corresponding to the reports of previous CT studies, our study showed a significant positive correlation between the FM length, width, and area.[7],[11] Similarly, this correlation was documented by Eboh[4] and Lyrtzis et al.[23] who evaluated the FM morphometry on skull bones. Eboh[4] and Abo El-Atta et al.[11] attributed this strong correlation to a homogenous growth rate of the cranium and proportional growth in the measurements of the FM.


The dimensions of the foramen magnum were sexually dimorphic. Their high overall accuracy (75%) in gender discrimination implies that they may collectively be used to positively estimate the gender of unknown skulls in Delta State Nigeria.

Strength of the study

The use of CT images in this study provided accurate measurements of the foramen magnum which were used to derive an equation for effective gender discrimination. The findings are therefore objective and more reliable in forensic investigations in our population.

Limitations of the study

A purposive sampling technique was adopted and only the available CT images from a single center were utilized. This was a small sample size inferring that our findings may not be generalized. In developing countries, the use of CT is limited to diagnostic purposes and is only available in a few selected hospitals. The forensic application of the findings of this study is therefore limited.


We would like to acknowledge Priscilla Ejiroghene and Dr. Jaiyeoba-Ojigho Jennifer Efe who assisted with data collection and data analysis, respectively.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.


1Gopal SK, Sushmitha S, Kumar M. Mastoid and magnum–Hidden key in forensics – A retrospective three-dimensional cone-beam computed tomographic study. Int J Forensic Odontol 2020;5:62-7.
2Agarwal HK, Setia PS, Pandey S. Virtual determination of sex: Estimating cut off value of digital metric traits of foramen magnum on three-dimensional computed tomography with receiver operating characteristic and logistic regression analysis. J Forensic Sci Med 2021;7:1-8.
3Diac MM, Hunea I, Girlescu N, Knieling A, Damian SI, Iliescu DB. Morphometry of the foramen magnum for sex estimation in Romanian adult population. BRAIN 2020;11:231-43.
4Eboh DE. Morphometric and morphological study of foramen magnum in dry adult skulls in a Southern Nigerian population. J Anat Sci 2014;5:48-53.
5Wani BA, Nazir N, Sheikh RA, Chalkoo AH, Jan T. Morphometric analysis of foramen magnum in the determination of sex using computed tomography. J Forensic Sci Med 2021;7:9-13.
6Kamath VG, Asif M, Shetty R, Avadhani R. Binary logistic regression analysis of foramen magnum dimensions for sex determination. Anat Res Int 2015;2015:459428.
7Verma P, Gupta N, Sameera Y, Faraz SA, Sharma P, Sharma B. Foramen magnum as determinant of sexual dimorphism in Sri Ganganagar population: A radiographic study. J Indian Acad Oral Med Radiol 2021;33:71-6.
8Vinutha SP, Suresh V, Shubha R. Discriminant function analysis of foramen magnum variables in south Indian population: A study of computerised tomographic images. Anat Res Int 2018;2018:1-8.
9Slima SR, Ragab E. Orbital and foramen magnum variables for sex determinations in Egyptians using computed tomographic images. Egypt J Forensic Sci Appli Toxicol 2020;20:73-83.
10Lopez-Capp TT, Rynn C, Wilkinson C, Paiva LA, Michel-Crosato E, Biazevic MG. Sexing the cranium from the foramen magnum using discriminant analysis in a Brazilian sample. Braz Dent J 2018;29:592-8.
11Abo El-Atta HM, Abdel-Rahman RH, El-Hawary G, Abo El-Al-Atta HM. Sexual dimorphism of foramen magnum: An Egyptian study. Egypt J Forensic Sci 2020;10:1-12.
12Akay G, Güngör K, Peker İ. Morphometric analysis of the foramen magnum using cone beam computed tomography. Turk J Med Sci 2017;47:1715-22.
13Silva RF, Picoli FF, Botelho TL, Resende RG, Franco A. Forensic identification of decomposed human body through comparison between ante-mortem and post-mortem CT images of frontal sinuses: Case report. Acta Stomatol Croat 2017;51:227-31.
14Loyal P, Ongeti K, Pulei A, Mandela P, Ogeng'o J. Gender related patterns in the shape and dimensions of the foramen magnum in an adult Kenyan population. Anat J Afr 2013;2:138-41.
15Gargi V, Prakash SM, Malik S, Nagaraju K, Goel S, Gupta S. Sexual dimorphism of foramen magnum between two different groups of Indian population: A cross-sectional cone-beam computed tomography study. J Forensic Sci Med 2018;4:150-5.
16Moodley M, Rennie C, Lazarus L, Satyapal KS. The morphometry and morphology of the foramen magnum in age and sex determination within the South African Black population utilizing computer tomography (CT) scans. Int J Morphol 2019;37:251-7.
17Lopez-Capp TT, de Paira LA, Buscath MY, Micheal-Crosato E, Biazevic MG. Sex estimation of Brazilian skulls using discriminant analysis of cranial measurements. Res Soc Dev 2021;10:1-16.
18Bartholdy BP, Sandoval E, Hoogland ML, Schrader SA. Getting Rid of dichotomous sex estimations: Why logistic regression should be preferred over discriminant function analysis. J Forensic Sci 2020;65:1685-91.
19Fleiss JL. Reliability of Measurement. The Design and Analysis of Clinical Experiments: NewYork. John Wiley & Sons, Inc; 1999. p. 1-32.
20Singh KC, Rai G, Rai R. Morphological variations of the foramen magnum in adult human dry skull in Eastern UP (India) population. Int J Med Res Prof 2017;3:205-8.
21Ilhan P, Kayhan B, Erturk M, Sengul G. Morphological analysis of occipital condyles and foramen magnum as a guide for lateral surgical approaches. MOJ Anat Physiol 2017;3:188-94.
22Lucena JD, Sanders JV, Brito HM, Cerqueira GS, Silva IB, Oliveira A. Morphometric analysis of the foramen magnum in dry human skulls in North-Eastern Brazil. J Morphol Sci 2019;36:97-104.
23Lyrtzis C, Piagkou M, Gkioka A, Anastasopoulos N, Apostolidis S, Natsis K. Foramen magnum, occipital condyles and hypoglossal canals morphometry: Anatomical study with clinical implications. Folia Morphol (Warsz) 2017;76:446-57.