Comparison of digital radiography and dual-sided reading of computed radiography system for imaging quality and radiation dose

Objective: To compare imaging quality and radiation dose on DR and dual-sided reading of CR system. Methods: The TRG of ALVIM statistics phantom was exposed by DR and dual-sided reading of CR with the same radiation dose, then the exposure factor and the entrance surface dose (ESD) were recorded. After that, the phantom was exposed by dual-sided reading of CR with various mAs ( 3.2, 4.0, 4.3, 4.8 and 5.2 mAs) and the same kV, and the ESD was recorded, too. The acquired images were read by three radiologists on the monitor of diagnosis work station with the same window width and window level, then the total points of the images were given with the method of a five-level confidence scale. ROC curves were drawn and probability Pdet was calculated. Then the data was analyzed statistically by analysis of variance (ANOVA). Results: The ROC analysis of bone substitute with diameter from 0.5 to 1.0 mm and muscle substitute with diameter from 0.9 to 2.0 mm showed A(z) values of 0.742 to 0.923 and 0.635 to 0.900 for DR system when the radiation dose was 137.5 μGy. The ROC analysis of bone substitute and muscle substitute showed A (z) values of 0.526 to 0.586 and 0.473 to 0.560 for dual-sided reading of CR at the same radiation dose. The probability Pdet discrepancy of bone substitute in two systems was from 0.216 to 0.337, with the highest of 0.337 in the diameter of 1.0 mm. The probability Pdet, discrepancy by muscle substitute in two systems was from 0.075 to 0.342, with the highest of 0.342 in the diameter of 2.0 mm. There was significant statistical difference between the two groups (F=35.30, P<0.01). When the exposure factor of dual-sided reading of CR system was 75 kV, 4.8 mAs, the ESD was 180.4 μGy, the ROC analysis in the bone substitute with diameter from 0.5 to 1.0 mm and muscle substitute with diameter from 0.9 to 2.0 mm showed A (z) values was the same. The probability Pdet, discrepancy of bone substitute in two systems was from -0.003 to 0.009. The probability Pdet, discrepancy of muscle substitute in two systems was from -0.005 to 0.008, and there was no statistical difference between the two groups (F=5.23, P> 0.05). When the probability Pdet of two systems was the same, the ESD was 180.4 μGy by dual-sided reading of CR system, and the ESD was 137.5 μGy by DR system. Conclusion: Detection with DR system was proved to be superior to dual-sided reading of CR system at the same radiation dose. When the image quality was similar, the radiation dose of DR system was less than the dual-sided reading of CR system. (authors)