In the present study, the amount of anterior extension of the maxillary sinus beyond the canine tooth long axis and the relationship between the root apex of the permanent maxillary canine and the maxillary sinus floor were evaluated in 300 CBCT images.
The proximity of the sinus floor and root apices of the maxillary teeth is of critical importance in several clinical procedures such as endodontic surgery, orthodontic treatment, and implant placement. CBCT scans, with dedicated 3D software allow the reconstruction of the image in three orthogonal planes precise morphometric measurements, and provision of accurate information for case selection, treatment planning, and avoidance of collateral damage during the surgery[17–19].
The relationship between the maxillary sinus floor and posterior teeth has already been evaluated in various studies [4, 10–16]. In contrast, few studies have addressed the relationship between the maxillary sinus floor and canine apex [4, 20–22]. AL-Qasab et al.  evaluated this relationship in a limited sample size (30 individuals) and in periapical radiographs. Ducommun et al. evaluated 83 CBCT images for the distance between the apices of the anterior teeth to the nasal floor or maxillary sinus, whichever was closer, and they did not report the results separately for each structure. Their study also included the anterior teeth with apical lesions. To the best of our knowledge, only one comparable study has been published recently which evaluated the relationships between the maxillary teeth (molars, premolars, and canine) and maxillary sinus floor in CBCT .
The result of the present study shows that most of the maxillary sinuses (68.83%) extended to the canine area which is almost equal to Zhang et al.’s finding (68.9%). In contrast, Kim et al.  and Kopecka et al.  reported much less percentage of canine area pneumatization by the maxillary sinus, (33% and 2.4%, respectively). Regarding the pneumatization of the incisor region, Zhang et al. reported a higher frequency (15.5%) in their study population compared to the present study (2.5%).
In this study, most of the maxillary sinuses (81.33 %) had bilateral symmetry with regard to the location of the anterior border in the jaw. This finding was in accordance with previous studies [2, 25, 26]. Comparably, Shahbazian et al.  reported the symmetric morphology of the maxillary sinus in 83% of their cases, and according to Hamdy et al., the average linear cranio-caudal, antero-posterior, and medio-lateral measurements were almost bilaterally-matched in all cases. Based on our results, bilateral symmetry was even more common in those maxillary sinuses which extended into the canine area. Further studies are required to confirm this result.
Despite the reported dissimilarity of the maxillary sinus volume between different genders[25, 29], the frequency of the maxillary sinus which extended to the anterior region was similar in both genders in the present study. This result is almost comparable to those reported by Jun et al.  who found no significant difference in the maxillary sinus volume between the males and females after the developmental period.
The frequency of the maxillary sinus which extended to the anterior region and also the mean AEMS beyond the canine long axis were significantly lower in older cases (40–50 years) compared with the other age groups. Additionally, in the vertical dimension, the MS-NF was found to be significantly higher in the older age group. These findings confirm those of Belgin et al. , Takahashi et al.  and Velasco-Torres et al.  who found decreasing maxillary sinus volume with increase in age. Jun et al.  also showed that the maxillary sinus increased in size until the completion of skeletal development in both sexes and then an age-related decrease occurred in its volume. Similarly, Ariji et al.  reported an increasing maxillary sinus volume up to 20 years of age and a decline process subsequently. Contrary to all these findings, Sahlstrand-Johnson et al.  reported that maxillary sinus volume was not related to age in their study sample.
Oishi et al. found a negative correlation between the distance of maxillary sinus floor to the root apices of all teeth (posterior teeth and canine) and age in CBCT images. The same result was found by AL Qasab et al.  regarding the distance between the canine apex and maxillary sinus floor in periapical radiographs. Similarly, we found that the distance between the maxillary sinus floor and canine apex increased with increase in age; however, the difference was not statistically significant.
As part of this study, it was found that when the maxillary sinus extended to the canine area, the canine apices, most commonly (84.99 %), was located below the sinus floor at more than 2 mm distance (type I). It was followed by types II and III relationship with much less percentage (8.96% and 6.05%, respectively). This order was in accordance with a recent study by Oishi et al.. They defined type 0 for those cases in which sinus floor did not appear above the root apices. Types I, II and III define the separate, in-contact and interlock relationship between the maxillary sinus floor and canine apices, respectively. They reported that whenever maxillary sinus appeared above the canine apices, type I was the most and type III was the least frequent relationship; this is similar to our findings.
Oishi et al. assessed the proximity of the maxillary canine and posterior teeth to the sinus floor in the standard sagittal and coronal planes. However, in the present study, the classification and measurements were done based on bucco-lingual cross-sections perpendicular to the dental arch of each side of the jaw. This corrected plane seems to be more sensible for evaluating the relationship and distance of the root apices and the sinus floor. In their study , the sample size weighted toward the female group (225) which was 3 times more than the male group (76). Moreover, the number of the younger subjects was twice that of the older age group. In the present study, we tried to select a similar number of male (154) and female (146) subjects and have a uniform subject distribution in different age groups to overcome the possible biases. We recommend further studies with lager sample size and wider age span in different population to evaluate the possible population-specific variation.