Association between subjective evaluation and physical parameters for radiographic images optimization.

PURPOSE: The purpose of this study was to develop a methodology to optimize computed radiographic techniques to image the skull, chest, and pelvis of a standard patient. METHODS: Optimization was performed by varying exposure levels with different tube voltages to generate images of an anthropomorphic phantom. Image quality was evaluated using visual grading analysis and measuring objective parameters such as the effective detective quantum efficiency and the contrast-to-noise ratio. Objective and subjective evaluations were compared to obtain an optimized technique for each anatomic region. RESULTS: Gold standard techniques provided a significant reduction in X-ray doses compared to the techniques used in our radiology service, without compromising diagnostic accuracy. They were chosen as follows 102 kVp/1.6 mAs for skull; 81 kVp/4.5 mAs for pelvis and 90 kVp/3.2 mAs for chest. CONCLUSION: There is a range of acceptable techniques that produce adequate images for diagnosis in computed radiography systems. This aspect allows the optimization process to be focused on the patient dose without compromising diagnostic capabilities. This process should be performed through association of quantitative and qualitative parameters, such as effective detective quantum efficiency, contrast-to-noise ratio, and visual grading analysis.

Quality and dose optimization in hand computed radiography.

The objective of the present study was to optimize a radiographic technique for hand examinations using a computed radiography (CR) system and demonstrate the potential for dose reductions compared with clinically established technique. An exposure index was generated from the optimized technique to guide operators when imaging hands. Homogeneous and anthropomorphic phantoms that simulated a patient's hand were imaged using a CR system at various tube voltages and current settings (40-55 kVp, 1.25-2.8 mAs), including those used in clinical routines (50 kVp, 2.0 mAs) to obtain an optimized chart. The homogeneous phantom was used to assess objective parameters that are associated with image quality, including the signal difference-to-noise ratio (SdNR), which is used to define a figure of merit (FOM) in the optimization process. The anthropomorphic phantom was used to subjectively evaluate image quality using Visual Grading Analysis (VGA) that was performed by three experienced radiologists. The technique that had the best VGA score and highest FOM was considered the gold standard (GS) in the present study. Image quality, dose and the exposure index that are currently used in the clinical routine for hand examinations in our institution were compared with the GS technique. The effective dose reduction was 67.0%. Good image quality was obtained for both techniques, although the exposure indices were 1.60 and 2.39 for the GS and clinical routine, respectively.