INTRODUCTION
The mental foramen (MF) is located in the anterolateral face of the body of the mandible, close to the inferior margin, generally between the first and second premolars (Olasoji et al., 2004; Al-Khateeb et al., 2007; El-Anwar et al., 2016; Fuentes et al., 2017). The visibility of the MF has been studied in panoramic radiographs in different populations. Ngeow et al. (2010) described a frequency of 77.8 % in a Malay population. Chkoura & El Wady (2013) analysed a Moroccan population sample and observed a frequency of 79 %. Fuentes et al. (2014) analysed the visibility of the MF in a Chilean population sample in conventional panoramic radiographs and reported a rate of 87.9 %, observing that detection of the MF is less frequent with advancing age and that it is more prevalent in men. The variations in this foramen have been studied in different populations in dry mandibles (Zografos & Mutzuri, 1989; Roopa et al., 2003), radiographic imaging techniques like CBCT (Naitoh et al., 2009a,b) and panoramic radiographs (Naitoh et al., 2011; Al-Shayyab et al., 2016; Srinivas et al., 2017); the reported prevalence varies between 3 % and 7 %.
The mandibular foramen (MBF) is located in the centre of the mandibular ramus, and is related with the mandibular lingula (Monnazzi et al., 2012). Its position has been studied in dry mandibles, with the conclusion that the MBF is symmetrical on the two sides; however the literature on the anatomical reference points used to locate it is still scarce (Marzola et al., 2005). The MBF may be accompanied by an accessory foramen, which may or may not contain the vasculo-nervous bundle (Narayana & Prashanthi, 2003; Das & Suri, 2004; Manikandhan et al., 2010). The position of this accessory foramen has been studied more frequently in dry mandibles (Thangavelu et al., 2012; Matveeva et al., 2018).
Digital panoramic radiographs offer good possibilities of detecting accessory foramina, due to the continuity and extent of the image. This examination, used for diagnostic purposes in dental practice, provides an image of sufficiently good quality to allow immediate viewing of the patient's whole maxillofacial area (Fuentes et al., 2014), as well as being quick and safe for users (Pyun et al., 2013). Discovering variations in the shape, size and position of the foramina has become an essential step for dental surgeons prior to any surgery, due to their intimate relation with the inferior alveolar nerve (Balcioglu et al., 2010; Gupta et al., 2015). Correct identification of these morphological variations is useful for avoiding injury during anaesthetic and surgical procedures, such as the placing of implants, periapical surgery, endodontic treatment and orthognathic surgery (Cillo & Stella, 2005; AlKhateeb et al.). The aim of this study was therefore to establish the prevalence of variations in the MF and MBF, and to carry out a morphometric analysis of a Chilean population using digital panoramic radiographs.
MATERIAL AND METHOD
A cross-sectional study was carried out, approved by the Scientific Ethics Committee (CEC) of Universidad de La Frontera (Folio no. 015/2014). The study analysed digital panoramic radiographs (ratio 1:1) taken using the standard technique with a PAX-400C orthopantomograph (VATECH, Korea, 2010), between January 2010 and December 2015, in the Dental Teaching Clinic of the Dental School, Universidad de La Frontera, Temuco, Chile. The measurements were made by two previously calibrated examiners and a third examiner, a dental surgeon specialising in radiology, who was responsible for taking a decision in doubtful cases.
A total of 1400 digital panoramic radiographs were available for analysis. The following exclusion criteria were applied: radiographs with inadequate image quality (distortion or alteration in the contrast), patients with teeth in intraosseous evolution (except third molars), presence of any pathology or development conditions which affected the visibility of the study area (tumours, cysts, fractures or malformations), isolated radio-opacities in edentate areas which may be associated with excessive ossification in a post operatory site, and radiographs showing evidence of orthognathic surgery.
Because it is impossible to tell which is the accessory foramen in cases of double or triple MF or MBF, all the foramina present were included in the analysis. To determine the morphometric measurements of double or triple foramina we used the ImageJ software (version 1.45), which allows the user to draw lines and to measure distances, áreas and angles of the various anatomical references observed in the radiograph. In each selected image we determined the area, width (greatest horizontal measurement) and height of the foramen (greatest vertical measurement) as shown in Figure 1. For statistical analysis we used the SigmaPlot 12.0 software, analysing the variables sex and age range (<30 years, >30 years) and relating them with the area, height (supero-inferior) and width (medio-lateral) of each foramen (double MF and double MBF).
RESULTS
In total, 927 radiographs were included in the study, after applied selection criteria, of which 327 belonged to men (35.27 %) and 600 to women (64.72 %). The patients' ages ranged between 18 and 86 years (mean age 36 ± 16.23). Only double MF and MBF were observed, no triple foramina were found.
Double mental foramen. We observed 24 cases (2.58 %) of double MF (Fig. 2A) among all the radiographs analysed. 58.33 % of cases (n=14) were men aged between 18 and 53 years, while 41.67 % of cases (n=10) were women aged between 22 and 54 years. The position of the double MF was in the right hemiarch in 54.17 % of cases (n=13) and in the left hemiarch in 45.83 % of cases (n=11). In men there was a preponderance of cases in the left hemiarch, 64.28 % of cases (n=9), while in women the preponderance was in the right hemiarch, 80 % of cases (n=8). The mean area of the double MF identified was 6.25 ± 3.96 mm2 (6.7 ± 3.66 mm2 for men and 5.85 ± 4.48 mm2 for women). Table I shows the mean area, width and height of the double MF identified, grouped by sex and age range. There were 6 groups, but a statistically significant difference was found only between the height and the patient's age (p=0.015), with the observation that the greater the subject's age, the smaller the height. No cases of bilateral double MF were found.
Double mandibular foramen. We observed 14 cases (1.51 %) of double MBF (Fig. 2B) among all the radiographs analysed. All cases were unilateral. 71.43 % of cases (n=10) were women aged between 18 and 55 years, while 28.57 % of cases (n=4) were men aged between 26 and 36 years. In women there was a preponderance of cases in the right hemiarch, 80 % of cases (n=8), while in men there was an even split between left and right hemiarch (50/50). The mean area of the double MBF identified was 6.55 ± 2.89 mm2 (6.79 ± 2.38 mm2 for men and 5.96 ± 3.12 mm2 for women). Table II shows the mean area, width and height of the double MBF identified, grouped by sex and age range into 6 groups. As with the MF, a statistically significant difference was found only between the height and the patient's age (p=0.019), with the observation that the greater the subject's age, the smaller the height of the double MBF.
Table I Distribution of the data by median, Q1 and Q3, and p value of six groups distributed by the relationship of the area, width and height of the double MF with the patient's sex and age range.

DISCUSSION
Various studies have reported the prevalence and characteristics of double MF and MBF, in different populations and using different instruments (Zografos & Mutzuri; Al-Khateeb et al.; Manikandhan et al.; Naitoh et al., 2011; Afkhami et al., 2013). The frequency of double MF has been studied in dry mandibles and in CBCT, with reported frequencies of around 7 % in Greek and Indian populations (Zografos & Mutzuri; Roopa et al.), and 3 % to 13 % in Asiatic populations (Katakami et al., 2008; Naitoh et al., 2009a,b; Olivera-Santos et al., 2011; Imada et al., 2014; Srinivas et al.). In our study of digital panoramic radiographs in a sample of Chilean patients, we report a prevalence of 2.58 %; this is lower than that described in the literature (around 5 %) (Naitoh et al., 2011; Al-Shayyab et al.; Srinivas et al.). The percentage difference in the prevalence may be due to the sample size (much larger in our case) and to ethnic origin (most studies of prevalence have been carried out in Asian populations). The predominant hemiarch has been studied in an Indian population, where the left side is described as more frequent. This is the case in Singh & Srivastava (2010), who found a prevalence of 5 % on the right side and 8 % on the left (Kalender et al., 2012). Similarly, Udhaya et al. (2013) described a frequency of 2.22 % on the right side and 3.33 % on the left. These data contrast with our results, which show a preponderance in the right hemiarch (1.4 %) over the left (1.18 %). By sex, the preponderance of double MF was in the left hemiarch in men, and the right hemiarch in women; however Al-Shayyab et al. found a preponderance on the right side in both men and women. The literature on the size of double MF is scarce. Some studies report the diameter, from which the calculated mean area is 1.34 mm (Katakami et al.; Naitoh et al., 2009a,b; Oliveira-Santos et al., 2011; Imada et al.; El-Anwar et al.); others report the area, for example Naitoh et al. (2011) who reported a mean area of 1.5 mm2 in a study based on 365 panoramic radiographs. In our study the mean value was 5.46 mm2 for both sexes. Of all the variables analysed, only the relationship between the height of the foramen and the patient's age presented statistically significant results (p=0.015), indicating that obliteration of the foramen in its supero-inferior dimension occurs with advancing age.
The MBF has been studied principally in dry mandibles. The reported prevalence of double MBF varies from 0.3 % to 32.36 % (Manikandhan et al.; Udhaya et al.; Shalini et al., 2016). Choi & Han (2014) analysed 446 CBCT records and described a prevalence of 1.35 %, similar to our results obtained from panoramic radiographs (1.51 %); there are few reports in the literature using this method. In the same study, which analysed three types of accessory mandibular canals, type 3 was a canal originating in another foramen than the MBF; the prevalence of double MBF reported by the authors was 0.06 % (Choi & Han). In our study, the analysis by sex showed that there was a higher prevalence of double MBF in women (71.42 %), and that the predominant location was in the right hemiarch. This agrees with Freire et al. (2012) results in a population in Brazil, but contrasts with the results of Naitoh et al. (2009a,b) who describe predominance in the left side in an Asian population. As with the double MF, we analysed the height in relation to age and observed that a diminution in the supero-inferior dimension of the foramen occurs in patients aged over 30 years. This agrees with the findings of Al-Khateeb et al. who describe that the invisibility of the foramen increases with advancing age and is accentuated in patients aged over 50 years.
Studying the anatomical variables in the foramina is important, because lack of knowledge of these variations could explain failures in nerve block in this zone during surgery and routine dental procedures using conventional anaesthetic techniques.
CONCLUSIONS
Accessory MF and MBF occur infrequently in the Chilean population; in our results only double foramina were found, which were more frequent in the MF. In both foramina the supero-inferior height diminished in patients aged over 30 years. Consideration of these anatomical variants is important for dental surgeons carrying out various clinical and surgical actions, therefore radiograph diagnosis by either conventional or high resolution methods is necessary to avoid possible complications.