Comparison of accuracy of the maxillary sinus area and dimensions for sex estimation lateral cephalograms of Iranian samples

Background

Sex estimation is a critical issue in forensic medicine. Since anthropologic evaluations on radiographs can contribute to more accurate sex estimation, this study measured the length and height, as well as area, of the maxillary sinus on lateral cephalometric radiography.

Methods and materials

In this descriptive-analytic study, 80 images obtained from 20 to 40-year-old individuals (40 males and 40 female) were selected. The height and anterior-posterior length of the maxillary sinus was performed using CATIA V5R20 (a software package for image analysis). Maxillary sinus area was determined using discriminant analysis in AutoCAD.

Findings

The mean maxillary sinus height was 40.48 mm in males and 38.7 mm in females. The mean maxillary sinus length was 40.31 mm in males and 37.31 mm in females. The area of the maxillary sinus was 1201.2 mm² in males and 1043.4 mm² in females. Two discriminant analysis models, one based on the height and length of the maxillary sinus and the other based on sinus area) were developed for sex estimation. The area under the receiver operating characteristic (ROC) curve was used for quality assessment of the fitted models and determination of their prediction ability.

Conclusion

The mean height, length and area of the maxillary sinus were significantly larger in males than in females (P < 0.05). While sex estimation can be performed based on either the area or a combination of the height and length of the maxillary sinus. The length/width perform better than area for sex estimation but, it is greatly overshadowed by the fact that both measures perform barely better than chance. Since classification accuracy is less than 80%, the method is not reliable and therefore not recommended for sex estimation.