AIM: To compare the diagnostic potential of direct digital images with conventional radiographs for the detection of simulated root canal voids. METHODOLOGY: The root canals of 80 extracted maxillary incisors with straight root canals were prepared and filled and the specimens divided into two groups: (i) a group of 40 teeth without voids; and, (ii) a group of 40 teeth with voids created in the root filling. Each root was imaged using the paralleling technique with a CCD-based digital system and F-speed film. Totally, there were 240 images: conventional radiographs (n = 80), digital images (n = 80) and digital zoomed images at 2x magnification (n = 80). These were interpreted for voids by a radiologist (without prior knowledge of the distribution of the voids). The reliability of the radiologist was confirmed using Cohen's Kappa Statistic, with the kappa value found to be >0.08. The sensitivity, specificity and predictive values were calculated. The differences in sensitivity and specificity between the imaging systems were then assessed using the Exact McNemar significance probability at the 95% significance level. RESULTS: The sensitivities (95% CI) of conventional radiographs, digital images and zoomed images were 0.68 (0.57-0.78), 0.83 (0.74-0.91) and 0.80 (0.71-0.89), respectively. The specificities of the conventional radiographs, digital and zoomed images were 0.80 (0.71-0.89), 0.83 (0.74-0.91) and 0.83 (0.74-0.91), respectively. The positive predictive values (95% CI) were 0.77 (0.68-0.86), 0.83 (0.74-0.91) and 0.82 (0.74-0.90), respectively. The negative predictive values (95% CI) were 0.71 (0.61-0.81), 0.83 (0.74-0.91) and 0.80 (0.72-0.89), respectively. There were no significant differences between the sensitivity and specificity of digital images and conventional radiographs, and of digital images and zoomed images during the detection of voids in root fillings (P > 0.05). CONCLUSIONS: Under the conditions of this laboratory study, the diagnostic performance of digital images and conventional radiographs, and between the digital images and digital zoomed images for simulated void detection in root fillings were not significantly different.
AIM: To compare the diagnostic potential of direct digital images with conventional radiographs for the detection of simulated root canal voids. METHODOLOGY: The root canals of 80 extracted maxillary incisors with straight root canals were prepared and filled and the specimens divided into two groups: (i) a group of 40 teeth without voids; and, (ii) a group of 40 teeth with voids created in the root filling. Each root was imaged using the paralleling technique with a CCD-based digital system and F-speed film. Totally, there were 240 images: conventional radiographs (n = 80), digital images (n = 80) and digital zoomed images at 2x magnification (n = 80). These were interpreted for voids by a radiologist (without prior knowledge of the distribution of the voids). The reliability of the radiologist was confirmed using Cohen's Kappa Statistic, with the kappa value found to be >0.08. The sensitivity, specificity and predictive values were calculated. The differences in sensitivity and specificity between the imaging systems were then assessed using the Exact McNemar significance probability at the 95% significance level. RESULTS: The sensitivities (95% CI) of conventional radiographs, digital images and zoomed images were 0.68 (0.57-0.78), 0.83 (0.74-0.91) and 0.80 (0.71-0.89), respectively. The specificities of the conventional radiographs, digital and zoomed images were 0.80 (0.71-0.89), 0.83 (0.74-0.91) and 0.83 (0.74-0.91), respectively. The positive predictive values (95% CI) were 0.77 (0.68-0.86), 0.83 (0.74-0.91) and 0.82 (0.74-0.90), respectively. The negative predictive values (95% CI) were 0.71 (0.61-0.81), 0.83 (0.74-0.91) and 0.80 (0.72-0.89), respectively. There were no significant differences between the sensitivity and specificity of digital images and conventional radiographs, and of digital images and zoomed images during the detection of voids in root fillings (P > 0.05). CONCLUSIONS: Under the conditions of this laboratory study, the diagnostic performance of digital images and conventional radiographs, and between the digital images and digital zoomed images for simulated void detection in root fillings were not significantly different.