Measurements and evaluation of the image noise power spectrum for computed radiography

Digital radiography has gained popularity in many areas of clinical practice. With this transition has come interest in advancing methodologies for image quality characterization. Modulation transfer function (MTF), noise power spectrum (NPS), and detective quantum efficiency (DQE) are widely accepted measurements of system performance. The NPS is one of the most common metrics describing the noise amplitude and texture observed in images obtained with a uniform field of radiation. Yet the comparability of the results is compromised by the different measurement parameters and evaluation methods. Because of the variety of methods used, it is difficult to compare previously published NPS results using different methodologies. In this paper, the following factors were evaluated for their effect on the measured NPS using the international electro-technical commission (IEC 62220-1)-defined RQA5 (additional Filter 21 mmAl, 72 kVp) radiographic techniques: impact of different exposure levels or radiation qualities, impact of the region of interest (ROI) and sub-ROI size, directional dependence, impact of overlapping versus non-overlapping sub-ROI size, and positional dependence for establishing the ROI. The results of Agfa CR (computed radiography, Toshiba, E8239X generator with Agfa CR 25 Corp.) showed that the shape of the normalized noise power spectrum (NNPS) curves decreased gradually as spatial frequency was increased. The various processing factors used in this study were found to have relatively little influence on the measured NPS. Our results suggest that the various processing factors could be used for computing the NPS with confidence. The results of this study could be used as a baseline to standardize the NPS processing by measuring NPS for different methodologies.