INTRODUCTION
The skull covering it is yet another one of a kind which not only protects it from the external environment but also allows it to communicate with the extracranial structures. The skull does not have the exact same morphology in every individual of human race. In fact, variations can be observed between two sides in the same individual. A wide variety of morphological variations in different races, geography, age, sex, etc., can be observed. One foramen of interest in view of its surgical importance is foramen ovale.
With the evolutionary history of the skull, it is evident that the foramen has been procured by the mammals only during the development and maturation of the organism and it is found to be absent in the mammalian embryo. Until the development of foramen ovale takes place, the Gasserian ganglion persists as an extracranial structure.
The foramen ovale is an obliquely placed, oval-shaped opening at the base of the skull. It is usually located in the greater wing of sphenoid bone, along the posterior margin of lateral pterygoid plate, antero-medial to foramen spinosum and lateral to the foramen lacerum.[ 1 ]
Foramen ovale connects the middle cranial fossa to the infratemporal fossa and transmits the mandibular nerve, accessory meningeal artery, lesser petrosal nerve, and emissary vein. The vertical axis of foramen corresponds to ventrolateral course of the mandibular nerve. The exact location and dimensions of foramina ovale play a vital role during certain diagnostic procedures such as electroencephalographic analysis, microvascular decompression by percutaneous trigeminal rhizotomy in trigeminal neuralgia, and fine-needle aspiration (percutaneous biopsy) through trans-facial approach in the perineural spread of tumor as in cavernous sinus tumors.[ 2 ] Insight on the topography and probable variations in the position allows us to avoid feasible injury to the mandibular branch of trigeminal nerve and other structures passing through foramen ovale during these procedures. Regional and ethnic discrepancies associated with the foramen prompted us to consider performing this study. This study thereby is of use to the neurosurgeons and other medical personnel to localize the foramen in the South Indian population.
SUBJECTS AND METHODS
Eighty foramina ovale of 40 dry adult South Indian human skulls were used to study the presence of possible variations in its morphometry. The sex of the skulls examined was not included in the study. Skulls with damage, fracture or deformity at the foramen ovale, and its surroundings were precluded from this study. The presence of the foramen ovale was confirmed bilaterally by observing the posterior part of the greater wing of the sphenoid bone. A thin copper wire was used to ascertain patency of foramina and rule out false passages. The nonmetric parameters studied bilaterally were as follows: (a) Shape of the foramen ovale and (b) presence of any accessory bony structure such as spur, tubercle, bony septa, and duplications. The metric parameters studied bilaterally were (a) Length of foramen ovale, (b) Width of foramen ovale, (c) area, (d) distance between zygomatic tubercle and the center of foramen ovale, D1 and (e) distance between midline and the center of foramen ovale, D2. Antero-posterior length and the Medio-lateral width of the foramina were measured along the oblique axis with the help of digital Vernier caliper with a precision of 0.01 mm [Figures 1 and 2 ].
Figure 1: Figure showing using probe to check the patency of foramen ovale
Figure 2: Figure showing measuring of dimensions of foramen ovale using Vernier caliper
Each dimension was measured thrice and the mean figure was recorded to prevent any error. From these obtained values, area “A” of the foramina ovale was calculated using the Radinsky’s formula: (p × L × B)/4.[ 3 ] Data analysis of the obtained values was performed statistically using the SPSS software version 11.5 for Windows, Chicago, Illinois, USA. The mean and standard deviation of each dimension was computed. Right and left differences were analyzed. A comparison was made of the means of the dimensions using the Student’s t -test. The relation between the variables was investigated by means of Pearson’s correlation coefficient (r), a probability (p) of <0.05 was considered statistically significant.
RESULTS
The foramen ovale was present bilaterally in all the skulls.
Nonmetric data analysis
Shape: In this study, the shape of the foramen ovale was “D” shaped in 35% while almond shaped in 21%, elongated oval in 15%, oval in 11%, irregular in 9%, longitudinal (slit-like) in 5%, and rounded in 4% of the foramina observed. The right-sided foramen ovale was almond in 30% while irregular in 17.5%, elongated oval in 15%, oval in 15%, “D” shaped in 12.5%, longitudinal (slit-like) in 7.5%, and round in 2.5% of the foramina observed. The left-sided foramen ovale was D shaped in 57.5% while elongated oval in 15%, almond in 12.5%, oval in 7.5%, rounded in 5%, longitudinal (slit-like) in 2.5%, and irregular in 0% of the foramina observed [Tables 1 -2 and Figures 3 -9 ]
Out of 40 skulls, the bilateral symmetric shape was observed in 15% skulls of which bilateral almond shaped foramen was seen in 50% while “D” shaped in 33.33% skulls and elongated oval in 16.66% skulls [Graph 1 ]
Accessory bony structure: Either or both the foramina in a skull had an accessory bony structure such as spur, tubercle, bony septa, duplication, and absent foramen spinosum in 72.5% skulls. Nevertheless 37.5% foramina showed no bony accessory structure bilaterally in skull. The margin of foramen of either or both of the foramina in the same skull had an accessory bony structure such as spur, tubercle, and bony septa present in 67.5% skulls and was smooth and regular in 32.5% skulls. A variation in either or both the foramina of a skull like spur was present in 47.5% skulls, bony septa in 7.5%, duplication in 32.5% skulls, and absent foramen spinosum was seen in 7.5% skulls [Graph 2 ]
Duplication: Duplication of foramen was seen in 32.5% skulls, bilateral duplication of foramen ovale was observed in 10% skull. Right-sided foramen duplication was seen in 10 out of 40 skulls (25%) and left-sided duplication in 7 out of 40 skulls (17.5%). Single duplication was seen in 21.25% foramina, of which 10 foramina were found to be on the right side and 7 were found to be left side. Multiple duplications in 2.5% foramina were also observed [Graph 3 , Figures 7 and 10 ]
Spur: 47.5% skulls showed the presence of one or more spur in the foraminal margin. Bilateral presence of spur in the foramen ovale margin was seen in 7.5% of skulls. Right-sided foramen with spur was 32.5% and left-sided foramen with spur was 22.5%. Single spur was present in 18 foramen (22.5%) and was observed to be more on the right side compared to the left. Double spur in the margin of the same foramen was found in 2.5% foramina and all of them belonged to the right-sided foramina. Triple spur in the margin of the same foramen was present in 2.5% of foramina with equal occurrence on both the sides [Figures 10 and 11 , Graphs 2 and 4 ]
Bony septa: 7.5% skulls presented with septate foramen. Foramina were found to be compartmented into two different compartments of unequal size by a bony septa perpendicular to its long axis. All of the septate foramen belonged to the right side. No skull presented with bilateral presence of septate foramen [Figures 12 , 13 and Graph 2 ]
Associated absent foramen spinosum: Foramen spinosum was found to be absent in 7.5% skulls. No skull presented with bilateral absence of foramen spinosum. Right-sided absence of foramen spinosum was observed in 2.5% and left-sided absence in 5% skulls [Figure 14 and Graph 2 ]. Asymmetry was present on the basis of accessory bony structure between the right and the left side foramen of the same skull in 100% cases.
Table 1: Percentage of various shapes of foramen ovale bilaterally
Table 2: Percentage of skulls showing bilateral symmetry
Figure 3: Figure showing almond-shaped foramen ovale
Figure 4: Figure showing elongated oval-shaped foramen ovale
Figure 5: Figure showing oval-shaped foramen ovale
Figure 6: Figure showing D-shaped foramen ovale
Figure 7: Figure showing irregular shaped foramen ovale with duplication
Figure 8: Figure showing slit-shaped foramen ovale
Figure 9: Figure showing round-shaped foramen ovale
Graph 1: Distribution of various shapes of foramen ovale
Graph 2: Analysis of the nonmetric data in foramen ovale
Graph 3: Number of foramen with duplication
Figure 10: Figure showing foramen ovale with a spur and duplication
Figure 11: Figure showing the presence of a spur in foramen ovale
Graph 4: Number of foramen with single, double, and triple spurs
Figure 12: Figure showing foramen ovale with septum
Figure 13: Figure showing the presence of a septum in foramen ovale
Figure 14: Figure showing absence of foramen spinosum with septate foramen ovale
Metric parameters analysis
Length and width
The asymmetry in length and width was present in 100% of skulls. The mean length and width of foramen ovale in skulls were 6.51 ± 1.24 mm and 3.66 ± 0.82 mm on right and 6.59 ± 1.33 mm and 3.75 ± 0.68 mm on the left, respectively [Table 3 ].
Table 3: Statistical analysis of foramen ovale
D1 and D2
The mean distance between zygomatic root and the center of foramen ovale (D1) was 30.11 ± 2.08 mm and 29.84 ± 2.09 mm on the right and left, respectively, with the mean distance between midline and the center of foramen ovale, (D2) was 23.25 ± 1.63 and 23.35 ± 1.42 mm on the right and left, respectively [Table 3 ].
Area
The mean area of foramen ovale calculated using Radinsky’s formula was found to be 19.08 ± 5.71 mm2 and 17.88 ± 5.85 mm2 on the right and the left, respectively [Table 3 ].
The Pearson’s correlation was calculated between length and area, width and area, D1 and D2, D1 and area, D2 and area, D1 and length, D2 and length, D1 and width and D2 and width. Of all the parameters included for the correlation study, a strong positive correlation was observed bilaterally between length and area (P value 0.00021 and < 0.00001 on right and left), width and area (P < 0.00001 on both sides), and D1 and D2 (P value 0.03582 and 0.005754 on right and left). Positive correlation was observed between D1 and length and D2 and length, but the relation was not significant. On the other hand, a negative insignificant correlation was seen between D1 and width and D2 and width. On performing the student t -test to compare the values between the right and left sides, the P value was observed to be more than 0.05 with all the parameters, thus making it insignificant.
DISCUSSION
Initially, in lower animals, the foramen ovale was found to be located on the spheno-parietal suture thus being a part of both the bones. With the evolutionary changes in the skull, the lateral growth of alisphenoids resulted in engulfment of the foramen within the greater wing of sphenoid in Homo sapiens while positional variations of foramen spinosum can be seen either on the spheno-squamous suture or on the bones on either side of the same. In the present study, the foramen ovale was entirely located medial to the sphenosquamous suture while variations were observed in the location of foramen spinosum in relation to the same suture.
Several cartilages fuse to form the cartilaginous base of the developing cranium, also called as chondro-cranium. With endochondral ossification, the bones in the base of the skull are formed. The ossification of bones in the base of skull takes place sequentially, the occipital bone being the first followed by body of sphenoid and ethmoid. The hypophseal cartilage forms around the developing hypophysis cerebri as pre- and postsphenoid centers which later fuses to form the body of the sphenoid.[ 4 ] The postsphenoid center also contributes to the formation of greater wing of sphenoid. The ossification center first appears for alisphenoid which on membranous ossification gives rise to the greater wing of sphenoid. At around 22 weeks, the foramen ovale can be seen surrounding the mandibular nerve as an unossified cartilage.
Endochondral and intramembranous ossification of basisphenoids in humans occurs at 14–17 weeks (body), orbitosphenoids (ala orbitalis/lesser wings) at the 16th week and alisphenoids (ala temporalis/greater wings) at the 15th week from the posterior to anterior aspect of the skull while ossification around the mandibular division of trigeminal nerve takes place later, to be visible as a distinct foramen at 22 weeks of gestation.
Asymmetrical size and shape of the foramina ovale of the same skull are present due to asymmetry in the emissary veins connecting cavernous venous sinus and pterygoid venous plexus on either side. These veins might be separated by dense fibrous connective tissues while transmitting through the foramen ovale and these fibrous bands might ossify occasionally forming duplications of foramen ovale or may present as lamina or bony septum dividing the foramen into compartments of unequal sizes.[ 1 ] Hence, the variations in the length and width of foramen with septum or duplication are reasonable.
The foramen ovale was present bilaterally in all skulls in the present study. Out of 40 skulls, the bilateral symmetric shape was seen in 15% skulls, of which bilateral almond shaped foramen was seen in 50% skulls while D-shaped in 33.33% skulls and elongated oval in 16.66% skulls.
In contrast to many other studies, the “D”-shaped foramen ovale was found to be most common shape observed in the present study [Table 4 ]. When compared with other Indian studies, the D-shaped was not a very common observation by many, but, the slit shaped was observed to be the least common by some studies.[ 1 , 5 , 6 ] While the shape that had the second highest incidence was oval and elongated oval in the present study, a study by Ray et al ., 2005 done on Nepalese skulls reported almond shape to be the second highest in record.[ 7 ] Unlike a study on Pakistani and Korean population which reported the irregular shape to be the least common, the present study shows the round-shaped foramen to be of least incidence [Table 5 ].[ 8 , 9 ] The present study also showed asymmetry of shape in 85% skulls. Despite all the variations in shape, bilaterally symmetrical foramina were observed only in 6 skulls accounting to a total of 15% only [Table 2 ].
Table 4: Comparison of shape of foramen ovale with other Indian studies
Table 5: Comparison of shape of foramen ovale with other International studies
The margin of foramen of either or both of the foramina in a same skull had an accessory bony structure such as spur, tubercle, and bony septa present in 67.5% skulls and was smooth and regular in 32.5% skulls. A variation like spur was present in 22 (27.5%) foramina in the present study in contrast to many other studies in India, which show a lesser incidence [Table 6 ].
Table 6: Comparison of accessory bony structures in the foramen ovale margin with other studies
Bony septa in 1.25% of foramen ovale was present in our study which corresponds to the study by Somesh et al. , 2011 showing 1.82% foramina having septa dividing the foramen into unequal compartments.[ 10 ] Bokhari et al . 2017 and Gupta and Rai 2013 found the presence of bony septa to be higher, i.e., 8.18% and 8.57%, respectively [Table 6 ]. [ 8 ]
The mean length of the right and left sides of the foramen was found to be similar to the studies by Poornima et al ., 2017 and Gupta and Rai, 2013 while it was found to be more pronounced in a study by Murugan and Saheb, 2014.[ 5 , 6 , 11 ] The mean width of the foramen in the present study was in concordance to the studies by Poornima et al ., 2017 and Gupta and Rai 2013, on the right side and by Murugan and Saheb, 2014 on the left side, while a study by Desai et al ., 2012 showed a much higher mean value on the left [Table 7 ].[ 5 , 6 , 11 , 12 ]
Table 7: Comparison of length and width with other Indian studies
On comparison with population from different nationalities, the length of the right-sided foramina was similar to the Pakistani, Nigerian, and Columbian population in contrast to the increased length in the Korean population.[ 8 , 13–15 ] On comparing the same dimension of the left foramina, Columbian, Nigerian, and Greek population had similar recordings in contrast to the Korean population.[ 3 , 13–15 ] The increase in the length of the foramina in the Korean population demonstrated the variation to be based on different races as well.[ 15 ] Similarly, the width was in correlation to the studies done in Nepalese, Greek, Turkish, Nigerian, and Columbian population while it was reported to be more in Germans on the right side.[ 3 , 7 , 13 , 14 , 16 , 17 ] However, in contrast, no wide range of difference was observed in the width of the foramina on the left side [Table 8 ]. In the present study, t -test was done and the P value thus obtained for the length and width of the foramina were found to be statistically insignificant, i.e., P > 0.05 (length = 0.571; width = 0.285). Despite the increased dimensions of foramen ovale in the Korean population, Hwang SH et al .,2005 reported the differences in the measurements to be statistically insignificant similar to the present study.[ 15 ]
Table 8: Comparison of length and width with other International studies
The area of the foramina was calculated using Radinsky’s formula: (π × L × B)/4 or (3.14 × L × B)/4 and was found to be 19.08 ± 5.71 mm2 and 17.88 ± 5.85 mm2 on the right and the left, respectively. Similarities in the calculated area were observed with most of the other Indian studies, American and Pakistani population. The mean area will slightly vary for the Korean population owing to the increased length in the foramina [Table 9 ].[ 3 ]
Table 9: Comparison of area of foramen ovale with other studies
The mean distance between zygomatic root and the center of foramen ovale, D1 was 30.11 ± 2.08 mm and 29.84 ± 2.09 mm on the right and left, respectively, in the present study. The mean distance between the midline of skull base and the center of foramen ovale, D2 was 23.25 ± 1.63 and 23.35 ± 1.42 mm on the right and left, respectively, in this study.
The foramina with duplication were relatively of smaller dimensions and the length was found to be increased in the foramina with bony septa extending mediolaterally, dividing it into unequal compartments and also when the foramen spinosum was absent in 3.75% foramina. The study done by Patil et al . 2013 on 102 foramina showed that the mean of distance from the center of foramen to the zygomatic arch (D1) was 32.58 ± 1.72 mm on right side and 32.75 ± 1.76 mm on left side, mean distance from the center of foramen to the mid-line of base of the skull (D2) was 25.83 ± 1.26 mm on the right side and 25.08 ± 1.31 mm on the left side [Table 10 ].[ 2 ]
Table 10: Comparison of D1 and D2 with other studies
The foramen ovale is a significant landmark to identify in various surgical procedures involving the middle cranial fossa. Its marking on the surface holds importance as it is necessary for various procedures like rhizotomy of the mandibular nerve in trigeminal neuralgia, for radiotherapy in reducing the size of tumors involving base of skull, to prevent injury to the structures passing through it in lesions involving cavernous sinus and so on. Trigeminal neuralgia is a common cause of excruciating facial pain where the patient’s personal, social, and professional life is tampered. Diseases such as sphenoidal mucocele, basilar encephalocele, carotid cavernous aneurysm, pituitary adenoma, tumors with metastatic lesions involving structures at base of skull can be dealt with through various approaches. However, their proximity to the foramen ovale during surgical procedures accentuates the increasing need for knowledge on its morphology and morphometry. The distance of the foramen ovale from the root of zygoma (D1) is of great significance as it confirms the zygomatic point, a point which is easily palpable topographically in front of the ear. The radiological markers can be placed over this point to gain an easy access to the foramen.
With the advent of key-hole surgeries, many approaches are available to access the foramen ovale which are more accurate and help prevent the associated morbidity when compared to open procedures. A substantial reduction in the manipulation of the brain is definitely a positive asset to the endoscopic procedures. Hartel’s trajetory which involves three anatomical landmarks acts as a guide to target the foramen ovale. Another endoscopic approach to the foramen by Chao Weiwei et al ., 2010 reports the distance between the nasal columella and foramen ovale to be 88.55 ± 5.00 mm in a radiological study and 89.75 ± 4.13 mm in an anatomical study. Another approach as reported by De Rosa et al ., 2019 elaborates about a superior orbital approach through endo-orbital and extra-orbital corridors created through the roof of orbit.[ 18 ] The study also reveals greater maneuverability through endo-orbital corridor and greater surgical freedom through extra-orbital corridor. A different approach through the molar planes and inter-eminence planes were superimposed and foramen ovale was identified in a study conducted by Zdilla et al . in 2016.[ 19 ] A lateral sub-labial approach to the foramen was described in another study where a submucosal sublabial incision was given and access was gained to the foramen by passing the probe under the zygomatic process of the maxilla just lateral to the lateral pterygoid plate Buzayed et al ., 2010.[ 20 ] On reaching the infratemporal fossa, the buccal and inferior alveolar nerves can be identified based on its course and relation to the lateral pterygoid muscle. On further tracing the buccal and inferior alveolar branches, the mandibular nerve can be identified and traced to the foramen ovale. According to a study by Alvernia et al ., 2010 the needle is inserted through an inverted pyramid which first passes till the parotid duct, then to the lateral pterygoid muscle and finally reaches the foramen ovale.[ 21 ] Although this technique is widely in use, care should be taken to avoid damage to branches of maxillary artery.
Although there are various approach routes available to access the foramen ovale, the percutaneous approach just below the zygomatic point seems to be involving a lesser rate of comorbidity. Care should be taken to prevent any injury to any of the branches of maxillary artery and accidental puncturing of pterygoid venous plexus as it can result in torrential internal bleed. The various branches of the mandibular nerve can be identified in the infratemporal fossa, and through one of the branches, Gasserian ganglion can be located on further tracing. The distance at which the foramen ovale is located from the zygomatic root is of great use commonly to treat trigeminal neuralgia. It is also helpful in placing the electrodes directly on the nerve during electroencephalography and also to correct mesial temporal lobe epilepsy. The distance between the center of foramen ovale and the midline of base of skull carries significance in case of any secondaries or primary tumors involving the base of skull, as it can result in shifting of the structures.
Financial support and sponsorship
This study was financially supported by Chettinad Hospital and Research Institute.
Conflicts of interest
There are no conflicts of interest.
Acknowledgment
We acknowledge the Chettinad Hospital and Research Institute nonteaching staff that helped us in the completion of this study.
REFERENCES
1. Mishra S, Sushobhana, Singh S, Singh R, Passey J. Morphometric and topographic study of foramen ovale in Indian skulls. Malays J Med Biol Res 2016;3:31–7.
2. Patil J, Kumar N, Mohandas Rao KG, Ravindra SS, Somayaji SN, Nayak BS, et al. The foramen ovale morphometry of sphenoid bone in South Indian population. J Clin Diagn Res 2013;7:2668–70.
3. Natsis K, Repousi E, Sofidis G, Piagkou M. The osseous structures in the infratemporal fossa:Foramen ovale, bony spurs, ossified ligaments and their contribution to the trigeminal neuralgia. Acta Neurochir (Wien) 2015;157:101–3.
4. Moore KL, Persaud TV, Torchia MG. The Developing Human Clinically Oriented Embryology. 8th ed. Philadelphia PA: Saunders/Elsevier; 2008. 347–48.
5. Poornima B, Sampada PK, Mallikarjun M, Santosh BS. Morphometric and morphological study of foramen ovale in dry adult human skull bones. Indian J Clin Anat Physiol 2017;4:59–62.
6. Murugan M, Saheb S. Morphometric and morphological study on foramen ovale. Int J Anat Res 2014;2:664–7.
7. Ray B, Gupta N, Ghose S. Anatomic variations of foramen ovale. Kathmandu Univ Med J (KUMJ) 2005;3:64–8.
8. Bokhari ZH, Munira M, Samee SM, Tafweez R. A morphometric study of foramen ovale in dried human skulls. Pak J Med Health Sci 2017;11:1661–5.
9. Kim DI, Kim HS. High resolution CT evaluation on the morphologic characteristics and variations of foramen ovale and adjacent foramina in the skull base. J Korean Radiol Soc 1995;33:43–8.
10. Somesh MS, Sridevi HB, Prabhu LV, Swamy MS, Krishnamurthy A, Murlimanju BV, et al. Amorphometric study of foramen ovale. Turk Neurosurg 2011;21:378–83.
11. Gupta N, Rai AL. Foramen ovale–morphometry and its surgical importance. IJMHS 2013;3:4–6.
12. Desai SD, Hussain SS, Muralidhar PS, Thomas ST, Mavishettar GF, Haseena S. Morphometric analysis of foramen ovale. J Pharm Sci Res 2012;4:1870–71.
13. Osunwoke EA, Mbadugha CC, Orish CN, Oghenemavwe EL, Ukah CJ. A morphometric study of foramen ovale and foramen spinosum of the human sphenoid bone in the southern Nigerian population. J Appl Bio Sci 2010;26:1631–5.
14. Muñoz NS, Rueda-Esteban RJ. Morphometric study of five constant skull base foramina in the Muisca population of the Tibanica anthropological collection of the Universidad de Los Andes. Int J Morphol 2016;34:1313–7.
15. Hwang SH, Lee MK, Park JW, Lee JE, Cho CW, Kim DJ. A morphometric analysis of the foramen ovale and the zygomatic points determined by a computed tomography in patients with idiopathic trigeminal neuralgia. J Korean Neurosurg Soc 2005;38:202–5.
16. Erbagci H, Kizilkan N, Sirikci A, Yigiter R, Aksamoglu M. Computed tomography based measurement of the dimensions of foramen ovale and rotundum in trigeminal neuralgia. Neurosciences (Riyadh) 2010;15:101–4.
17. Berlis A, Putz R, Schumacher M. Direct and CT measurements of canals and foramina of the skull base. Br J Radiol 1992;65:653–61.
18. De Rosa A, Pineda J, Cavallo LM, Di Somma A, Romano A, Topczewski TE, et al. Endoscopic endo-and extra-orbital corridors for spheno-orbital region:Anatomic study with illustrative case. Actaneurochirurgica 2019;161:1633–46.
19. Zdilla MJ, Hatfield SA, Mangus KR. Angular relationship between the foramen ovale and the trigeminal impression:Percutaneous cannulation trajectories for trigeminal neuralgia. J Craniofac Surg 2016;27:2177–80.
20. Buzayed B, Tanriover N, Canbaz B, Akar Z, Gazioglu N. Lateral sublabial endoscopic approach to foramen ovale:A novel endoscopic technique to access infratemporal fossa. J Craniofac Surg 2010;21:1241–5.
21. Alvernia JE, Sindou MP, Dang ND, Maley JH, Mertens P. Percutaneous approach to the foramen ovale:An anatomical study of the extracranial trajectory with the incorrect trajectories to be avoided. Acta Neurochir (Wien) 2010;152:1043–53.
