The aim of this study was to characterize the effect of an image processing algorithm (FineView) on both quantitative image quality parameters and the threshold contrast detail response of the GE Senographe DS full-field digital mammography system. The system was characterized using signal transfer property, pre-sampling modulation transfer function (MTF), normalized noise power spectrum (NNPS) and detective quantum efficiency (DQE) of the system. An algorithmic modulation transfer function (MTF(a)) was calculated from images acquired at a reduced detector air kerma (DAK) and with the FineView algorithm enabled. Two sets of beam conditions were used: Mo/Mo/28 kV and Rh/Rh/29 kV, both with 2 mm added Al filtration at the x-ray tube. Images were acquired with and without FineView at four DAK levels from 14 to 378 µGy. The threshold contrast detail response was assessed using the CDMAM contrast-detail test object which was imaged under standard clinical conditions with and without FineView at three DAK levels from 24 to 243 µGy. The images were scored by both human observers and by automated scoring software. Results indicated an improvement of up to 125% at 5 mm⁻¹ in MTF(a) when FineView was activated, particularly at high DAK levels. A corresponding increase of up to 425% at 5 mm⁻¹ was also seen in the NNPS, again with the same DAK dependence. FineView did not influence DQE, an indication that the signal to noise ratio transfer of the system remained unchanged. FineView did not affect the threshold contrast detectability of the system, a result that is consistent with the DQE results.

The purpose of the work was to test if effective detective quantum efficiency (eDQE) could be useful for optimisation of radiographic factors for computed radiography (CR) for adult chest examinations. The eDQE was therefore measured across a range of kilovoltage, with and without an anti-scatter grid. The modulation transfer function, noise power spectra, transmission factor and scatter fraction were measured with a phantom made of sheets of Aluminum and Acrylic. The entrance air kerma was selected to give an effective dose of 4.9 μSv. The effective noise equivalent quanta (eNEQ) is introduced in this work. eNEQ can be considered equal to the number of X-ray quanta equivalent in the image corrected for the amount of scatter and the blurring processes. The eNEQ was then normalised to account for slight differences in the effective dose (eNEQ ED ). The peak eNEQ ED was largest at 80 kV and 100 kV with no grid and with grid respectively. At each kilovoltage, the eNEQ ED and eDQE were between 10% and 70% larger when the grid was not used. The results show that 80 kV without grid is the most suitable exposure conditions for CR in chest. This is consistent with clinical practice in the UK and previous publications recommending a low kV technique for CR for average sized adult chest imaging.