NOISE POWER SPECTRA AND LOW CONTRAST RESOLUTION IN MULTIPLANAR REFORMATIONS

摘要

Low contrast imaging is of vital clinical importance and thus an important aspect of CT image quality. For example, soft plaque imaging, ground glass lung opacities, and soft tissue differentiation all depend upon excellent low contrast resolution. A primary factor in determining low contrast resolution is noise. Noise is generally characterized via the standard deviation. However, standard deviation reflects only on the magnitude of noise and not on the spatial frequency distribution of the noise. The spatial frequency distribution of noise can significantly impact SNR, depending on the frequency distribution of the object signal, and is thus important to noise characterization. In contrast to standard deviation, the noise power spectrum (NPS) yields both the noise magnitude and spatial frequency distribution.