RESUMEN
Bone density is frequently used in medical diagnosis and research. The current methods for determining bone density are expensive and not easily available in dental clinics. The aim of this study was to design and evaluate the accuracy of a digital method for hard tissue densitometry which could be applied on personal computers. An aluminum step wedge was constructed. 50 E-speed Kodak films were exposed. Exposure time varied from 0.05[s] to 0.5[s] with 0.05[s] interval. Films were developed with automatic developer and fixer and digitized with 1240U photo Epson scanner. Images were cropped at 10x10mm size with Microsoft Office Picture Manager. By running the algorithm designed in MATLAB software, the mean pixel value of pictures was calculated. Finding of this study showed that by increasing the exposure time, the mean pixel value was decreased and at step 12,[a] significant discrimination was seen between the two subsequent times [P<0.001]. By increasing the thickness of object, algorithm could define the density changes from step 4 in 0.3[s] and 5 in 0.5[s], and it could determine the differences in the mean pixel value between the same steps of 0.3[s] and 0.5[s] from step 4. By increasing the object thickness and exposure time, the accuracy of the algorithm for recognizing changes in density was increased. This software was able to determine the radiographic density changes of aluminum step wedge with at least 4mm thickness at exposure time of 0.3[s] and 5 mm at 0.5[s]