抄録
Purpose: We have developed and assess INCUS (inhomogeneous noise correction combined with uniform filter and sensitivity map), a novel technique to correct spatially inhomogeneous noise in surface-coil-based standard or parallel imaging in magnetic resonance (MR) imaging.
Materials and Methods: We employed a weighted summation of 2 images uniformly filtered with both a strong filter and a weak or no filter to achieve spatially nonuniform filtering by utilizing a coil-sensitivity and/or noise map to give greater weight to the strongly denoised components of the region with lower signal-to-noise ratio (SNR). We compared the image quality and difference between INCUS and standard uniform filter techniques employing several types of linear or nonlinear filter in abdominal or diffusion-weighted brain images each for 1 volunteer acquired on a 1.5-tesla magnetic resonance (MR) imager with parallel imaging after adding simulated noise.
Results: The INCUS technique inherently reduced inhomogeneous noise on multiple surface-coil imaging and minimized blur obtained with every filter type. Overall errors were smaller for the INCUS filter than the uniform filter when either of these was used in combination with any of the other filters.
Conclusion: The INCUS technique provides a straightforward, higher degree of freedom, and computationally feasible implementation of a general purpose de-noising filter for surface coil-based imaging including parallel imaging in commercial MR imaging.
