Development and validity evaluation of a screening program for Korean osteoporotic women using mandibular cortical width on panoramic radiograph

Abstract

Purpose The purpose of this study was to develop and evaluate a screening program for Korean osteoporotic women using mandibular cortical width (MCW), which has been used for bone mineral density (BMD) assessment on panoramic radiographs.

Materials and Methods A computer-based program was developed in order to measure the MCW automatically on panoramic radiographs. To identify changes in the MCW value according to the head position, panoramic radiographs were taken with 5 different vertical angulations using a head phantom. Normality test was performed among the MCW values measured from the panoramic radiographs of 250 young women. After a normal distribution was confirmed, the threshold corresponding to −2.5 standard deviations (SDs) was determined. The determined value was applied to the panoramic radiographs of 70 female subjects with a known femur BMD value, and the sensitivity, specificity, and accuracy were calculated. Additionally, a cut-off value for screening was obtained from a receiver operating characteristic (ROC) based on the data from these 70 female subjects.

Results There was no statistically significant difference in the MCW with a change of vertical angle of the head phantom (Kruskal-Wallis test, P=0.406). Analysis of the MCW in the panoramic radiographs of young females showed a normal distribution (P=0.074). The threshold value corresponding to −2.5 SD was 2.46 mm. When this value was applied to patients with a known femur BMD value, the sensitivity, specificity, and accuracy were 60.0%, 96.7%, and 91.4%, respectively. The area under the ROC curve was 0.947 (95% confidence interval 0.894-0.999, P=0.000). The cut-off value obtained from the ROC curve was 3.32 mm.

Conclusion In conclusion, it is suggested that the developed computer-based screening program for osteoporosis may have a validity in Korean women. If the cut-off value would be obtained from a big-data study, it could be used to screen for osteoporosis in patients who have undergone panoramic radiographic imaging in dental clinics.