Methods: Series of digital radiographic images were made of a dry mandible, using a Sidexis sensor and a Digora system. The X-ray beam was varied in a horizontal plane in steps of 3 degrees from –15 to +15 degrees with respect to the direction perpendicular to the mandible. Subtraction images were made combining images made with the same sensor and with the combination of sensor systems, without and with use of the reconstruction program. The standard deviation of the pixel grey values (GLSD) in the subtraction images was determined within a representative rectangular area, and used as the measure of similarity. Differences were considered statistically significant when p < 0,05.
Results: The reproducibility of the reconstruction procedure for both sensor systems was better than 95%. The GLSD of the reconstructed images was significantly lower than those of the untreated images (Sidexis p = 0,004, Digora p < 0,0001). The differences between reconstructed Sidexis and reconstructed Digora subtractions were not significant. Comparison of reconstructed subtractions of Sidexis and of Digora images and of subtractions obtained with the combination of sensor systems did not reveal significant differences.
Conclusions: The program is able to correct adequately the differences between pairs of images acquired from the Sidexis and Digora sensor systems. Research is continued to confirm the diagnostic benefit of this approach in a clinical environment.