Measurement accuracy: a comparison of two intra-oral digital radiographic systems, RadioVisiography-S and FlashDent, with analog film.

Abstract

To compare the accuracy of measurement algorithms incorporated into the software provided with two CCD-based digital radiographic systems and compare assessments using intra-oral film. A test object with three radiopaque reference points was imaged using Ektaspeed intra-oral film, both with and without an overlaid 1 mm reference grid, the RVG-S, the measurement software of which overlays a 2 mm grid, and the FlashDent, which uses a mouse-driven cursor to estimate distances. Geometric factors, including source-to-object and receptor-to-object distances, vertical and horizontal cone angulation, and vertical angulation of receptor, were successively altered. Six viewers estimated vertical, horizontal and diagonal distances for each image. Increased source-to-object distance resulted in a 5-10% reduction in image dimensions, whereas increased receptor-to-object distance produced a magnification of 15-20% with all techniques. Increased vertical and horizontal cone angulation resulted in slight increases in magnification (0-5%) whereas greater receptor angulation resulted in noticeable magnification under all test conditions (10-15%). Estimated values for vertical, horizontal and diagonal measurements were approximately 5-10% higher than the actual dimensions, irrespective of the technique used. The FlashDent mouse-driven cursor system and direct measurement of distances from conventional film were more accurate in horizontal and diagonal dimension estimates than using an overlaid grid. Intra-observer and inter-observer differences were found for specific situations. The mouse-driven computerized measurement cursor and intra-oral film assessment using a ruler provided the most accurate, reliable and consistent dimensional measurements.