(Sunday, 7/12/2020) [Eastern Time (GMT-4)]
Room: AAPM ePoster Library
Purpose: To investigate the NPS as a quality control constancy test for digital x-ray systems.
Methods: Uniform images were acquired under different conditions representing deviations from ideal performance using a digital x-ray system (Definium, GE Healthcare). Those included variations in tube potential, focal spot (resolution change) and defective pixels. The linearity of detector response was verified. Normalized NPS was calculated using the methodology of the international electrotechnical commission (IEC 62220-1). The NPS was computed for images with kV deviation (±5kV from nominal), resolution deviation (obtained by changing he focal spot), images with and without synthetically generated pixel defects, images with mismatched grid ratio and images with varying amount of residual signal. The NPS was also decomposed into its fixed pattern, quantum and electronic components to determine if the change in the NPS components reflected the expected change in the quantum, fixed pattern or electronic noise components.
Results: Measurable NPS differences could be detected for a deviation of 1 kV from the nominal value. The NPS also showed measurable difference when the focal spot size was changed to mimic a change in resolution, and when a small number of defective pixels were introduced in the image (0.02% of total pixels). The NPS clearly showed a spike in the noise spectrum at the spatial resolution of the mismatched grid septa. The NPS decomposition showed noticeable changes in fixed pattern and quantum components for kV and spatial resolution deviations and mismatched grid. With defective pixels, the fixed pattern component changed while the change in quantum component was indistinguishable. The NPS was not sensitive for residual image signal on this system.
Conclusion: This study suggests NPS is sensitive to deviations in system performance. its decomposition may help identify the source of deviation. The NPS is useful as quality control constancy test.
Funding Support, Disclosures, and Conflict of Interest: The funding for this study was received from CancerCare Manitoba Foundation.
Keywords
Noise Power Spectrum, Quality Control, Diagnostic Radiology
Taxonomy
IM- X-Ray: Quality Control and Image Quality Assessment
Contact Email
