Image quality and radiation dose comparison of a computed radiography system and an amorphous silicon flat panel system in paediatric radiography: a phantom study

摘要

This purpose of this work was to investigate the patient radiation doses and image quality of a Philips/Agfa computed radiographic (CR) system and a Philips indirect-capture digital radiographic (DR) system in a paediatric setting. A CDRAD digital radiographic contrast-detail phantom was used to assess radiographic image quality. Perspex slabs of three different thicknesses (6, 11 and 16 cm) were used to simulate paediatric patients of three arbitrary ages. These phantoms, in conjunction with the CDRAD digital radiographic contrast-detail phantom, were imaged under three different conditions. The CDRAD Analyser software package was used to assess the quality of each image. The first experiment conducted was a comparison of the two systems under standard conditions, with beam filtration of aluminium and copper, as recommended in European Guidelines on Quality Criteria for Diagnostic Radiographic Images in Paediatrics (European Commission 1996b). Image quality was compared for each phantom size at three doses with the same entrance exposure used for both systems. A visual comparison of the resulting contrast detail curves showed the DR system generally outperformed the CR system, especially at the lowest two doses. A chi-square analysis of the targets detected generally confirmed this visual impression. The second experiment performed was to compare the two systems under the conditions used in routine clinical practice at PMH. As a result of additional beam filtration not generally being employed, the image quality of the CR system was similar to the DR system for the two smaller phantom sizes but with a major dose cost - effective doses higher by between 38% and 100%. A chi-square analysis of the targets detected showed the CR system to be significantly better than the DR system at two of three doses for the thinnest phantom and no significant difference at any doses for the intermediate phantom size. For the largest phantom size, additional filtration - although different - was used for the CR and DR systems and so the X-ray beam spectra were more similar. Consequently, the results for this phantom size reflected those from the experiment conducted under standard conditions, ie the effective doses for both systems were similar and the image quality of the DR system superior. The chi-square analysis s howed the DR system to be significantly better than the CR at all three dose levels. A third experiment was undertaken to compare doses between the two systems at 'equal' image quality. The CDRAD Analyser software specific image quality parameter, IQFinv, was held constant for both systems. The entrance exposures required to achieve this image quality were measured and then converted to effective doses using the dose calculation software package PCXMC 1.5. The DR system offered effective dose savings of between 28 and 42% for the three phantom sizes. Overall, this work suggests that a Philips flat-panel system is superior to an Agfa CR system in paediatric radiography. This result generally reflects the findings of other authors who have conducted similar studies in adult patient settings.