Evaluating the effect of reduced entrance surface dose on neonatal chest imaging using subjective image quality evaluation

Abstract

IntroductionAs hospitalized neonates are radiosensitive and often require numerous X-rays, institutions should investigate the optimal beam parameter combinations to deliver diagnostically acceptable quality images at the lowest possible entrance surface dose (ESD). Using a subjective approach, this study evaluated the effect of different beam parameter combinations on image quality. Methods and materialsFive rabbits simulated the neonatal chest. The ESD was reduced using a variation of voltage, tube current and filtration. Eight radiology registrars, blinded to the dose parameter information, ranked the digital X-ray images of three anatomical regions from best to worst using a variation of the multiple rank order method. T-tests compared the average values, obtained from the scores assigned by each observer to images acquired at different ESDs, for each region. The calculated intraclass correlation coefficient (ICC) assessed observer agreement. ResultsResults showed that a 64% dose reduction was achievable by altering the beam parameters. Large variation among the observers was confirmed by an ICC value <0.5. A 95% confidence interval could not conclude that different ESD values, resulting in a 50–77% dose reduction compared to current practice, would result in different overall observed image quality. This was noted for all three regions indicating that no preference existed towards an image acquired with a specific beam parameter combination. ConclusionsThe large variation in observers’ opinion of acceptable image quality emphasizes the importance of subjective image quality evaluation in the clinical environment. The rabbit phantom and multiple rank order method are considered appropriate for these evaluations.