Image quality, threshold contrast and mean glandular dose in CR mammography.

In many countries, computed radiography (CR) systems represent the majority of equipment used in digital mammography. This study presents a method for optimizing image quality and dose in CR mammography of patients with breast thicknesses between 45 and 75 mm. Initially, clinical images of 67 patients (group 1) were analyzed by three experienced radiologists, reporting about anatomical structures, noise and contrast in low and high pixel value areas, and image sharpness and contrast. Exposure parameters (kV, mAs and target/filter combination) used in the examinations of these patients were reproduced to determine the contrast-to-noise ratio (CNR) and mean glandular dose (MGD). The parameters were also used to radiograph a CDMAM (version 3.4) phantom (Artinis Medical Systems, The Netherlands) for image threshold contrast evaluation. After that, different breast thicknesses were simulated with polymethylmethacrylate layers and various sets of exposure parameters were used in order to determine optimal radiographic parameters. For each simulated breast thickness, optimal beam quality was defined as giving a target CNR to reach the threshold contrast of CDMAM images for acceptable MGD. These results were used for adjustments in the automatic exposure control (AEC) by the maintenance team. Using optimized exposure parameters, clinical images of 63 patients (group 2) were evaluated as described above. Threshold contrast, CNR and MGD for such exposure parameters were also determined. Results showed that the proposed optimization method was effective for all breast thicknesses studied in phantoms. The best result was found for breasts of 75 mm. While in group 1 there was no detection of the 0.1 mm critical diameter detail with threshold contrast below 23%, after the optimization, detection occurred in 47.6% of the images. There was also an average MGD reduction of 7.5%. The clinical image quality criteria were attended in 91.7% for all breast thicknesses evaluated in both patient groups. Finally, this study also concluded that the use of the AEC of the x-ray unit based on the constant dose to the detector may bring some difficulties to CR systems to operate under optimal conditions. More studies must be performed, so that the compatibility between systems and optimization methodologies can be evaluated, as well as this optimization method. Most methods are developed for phantoms, so comparative studies including clinical images must be developed.

A new mammography dosimetric phantom.

Breast phantoms produced with tissue-equivalent materials are used in an attempt to simulate glandular and adipose tissues, in terms of X-ray attenuation and density. In this work, a set of breast tissue-equivalent phantoms (BTE phantoms) with semicircular shapes of different thicknesses and compositions were produced. Such phantoms may be used in the measurement of the incident air kerma (K(i)) and the mean glandular dose (D(G)) delivered to patients undergoing mammography. To characterise the materials used to produce the phantoms, a series of 17-keV X-ray attenuation coefficient measurements were performed. The carbon-nitrogen-hydrogen elemental composition and the densities of the tissue-equivalent materials were also determined and compared with values available in the literature. Linear attenuation coefficients of 0.724 and 0.923 cm(-1) were determined, respectively, for adipose and glandular tissues. Such values agree with data available in the literature. On the basis of the results obtained in this work, it is suggested that BTE phantoms are used instead of polymethyl methacrylate phantoms to select exposure parameters (kV, mAs and target/filter combination) specific for breast glandularities from 0 to 50 % in the optimisation of doses in mammography.