This study was a digital retrospective study. For this study, 600 digital OPGs of patients, both male and female aged 21 to 50 years, fulfilling the exclusion and inclusion criteria, were randomly selected from the archives of the department of oral medicine and radiology. The scans were anonymized before the evaluation of the selected parameters. The scans included in the study were of completely dentulous mandibles with proper positioning. The scans that were not included in the study were those where gross deformity of the maxilla-mandibular structures could be seen with any artifacts on the radiograph, any lesions in the mandibular arch that were radiolucent or radiopaque, premolars that were missing, mixed dentition, trauma history and/or ongoing or completed trauma treatment, and radiographic evidence of temporomandibular joint (TMJ) problems. There were two providers involved in collecting the measurements. One was a subject expert with more than 20 years of experience with OPG and its interpretations. The other was a postgraduate student with more than one year of experience in OPG and its interpretation.
A pilot study of 30 scans was done by both observers. The interobserver reliability was assessed through kappa statistics for the degree of agreement. From the overall kappa statistics value, we found out that an almost perfect agreement was achieved between the two providers overcoming any chance of bias and good generalizability. Calibration was then done for both observers, and then the remaining research was completed. Also to eliminate the measurement and provider bias, the analysis of the same anonymized scan was done by both observers twice each with a gap of 15 days.
A total of seven measurements (in mm) were performed on the digital radiographic image (Table1 ).
The method of analysis of various parameters was as follows on an OPG (Figure1 ).
These parameters were most stable and easy to identify on the mandible. The most convex point on the inferior border of the mandible in the posterior region is the lowest point on the bone. Hence, they were chosen for this study, which had a large sample size of 600 OPGs, after consulting a biostatistician and considering previous studies. These parameters were also considered in a previous study by Indira et al. [2 (link)] and Mostafa and El-Fotouh [13 ] that had a smaller sample size (<100).
The measurements were noted in tabular form. Each gender group was then split into three groups based on age to find out how the parameters of the mandible changed with age. The obtained data for categorical variables was shown as n% of cases, while the data for continuous variables was shown as mean and standard deviation (SD). The statistical intergroup comparison was estimated using the independent sample t-test or unpaired t-test. The statistical comparison of the intergroup distribution of means of continuous variables was also tested using the analysis of variance (ANOVA) procedure. For paired comparisons of means of continuous variables, the paired t-test was used. The linear discriminant function analysis was carried out to obtain a linear combination of various measurements that characterize the two classes of gender. Any underlying normality assumption was evaluated before applying the study variables for the t-test and ANOVA, and P-values < 0.05 were considered to be statistically significant. IBM SPSS Statistics for Windows, Version 21.0. (IBM Corp., Armonk, NY, USA) was used to figure out gender in the statistical analysis.
A pilot study of 30 scans was done by both observers. The interobserver reliability was assessed through kappa statistics for the degree of agreement. From the overall kappa statistics value, we found out that an almost perfect agreement was achieved between the two providers overcoming any chance of bias and good generalizability. Calibration was then done for both observers, and then the remaining research was completed. Also to eliminate the measurement and provider bias, the analysis of the same anonymized scan was done by both observers twice each with a gap of 15 days.
A total of seven measurements (in mm) were performed on the digital radiographic image (Table
The method of analysis of various parameters was as follows on an OPG (Figure
These parameters were most stable and easy to identify on the mandible. The most convex point on the inferior border of the mandible in the posterior region is the lowest point on the bone. Hence, they were chosen for this study, which had a large sample size of 600 OPGs, after consulting a biostatistician and considering previous studies. These parameters were also considered in a previous study by Indira et al. [2 (link)] and Mostafa and El-Fotouh [13 ] that had a smaller sample size (<100).
The measurements were noted in tabular form. Each gender group was then split into three groups based on age to find out how the parameters of the mandible changed with age. The obtained data for categorical variables was shown as n% of cases, while the data for continuous variables was shown as mean and standard deviation (SD). The statistical intergroup comparison was estimated using the independent sample t-test or unpaired t-test. The statistical comparison of the intergroup distribution of means of continuous variables was also tested using the analysis of variance (ANOVA) procedure. For paired comparisons of means of continuous variables, the paired t-test was used. The linear discriminant function analysis was carried out to obtain a linear combination of various measurements that characterize the two classes of gender. Any underlying normality assumption was evaluated before applying the study variables for the t-test and ANOVA, and P-values < 0.05 were considered to be statistically significant. IBM SPSS Statistics for Windows, Version 21.0. (IBM Corp., Armonk, NY, USA) was used to figure out gender in the statistical analysis.
Free full text: Click here
