Evaluation of Pixel Values from Digitized Radiographic Image

Objectives: Flatbed scanners with transparency option could be used as well as digital cameras to digitize radiographs. This study evaluated their performance concerning the pixel value of the final image. Methods: A ten-stepped exposure on occlusal film (Insight, Kodak, Rochester) was obtained by a sensitometer (Dual-Flashing, Nuclear Associates, USA). The optical densities were measured (Little Genius, Nuclear Associates) and turned into ideal pixel values by an algorithmic calculation. Using the same capture parameters, the image was digitized repeatedly ten times by the scanner HP 5370C. Results: The digitized images had mean pixel values higher than the ideal pixels for each one of the ten steps (T-student, p<0.01). The pixel value variation was around 10% from step 1 to 5 (0.25 to 0.28 OD); 9% for step 6 (0.32 OD); 7% for step 7 (0.41 OD); 5% for step 8 (0.65 OD); 3% for step 9 (1.15 OD) and 10 (2.02 OD). Conclusions: Pixel values of a digitized image could be modified according to the equipment used. The pixel value variation was higher at the less dense steps. Further studies are needed to extend these results to other models of scanners and digital cameras. (Supported by CAPES).