ABSTRACT
The main source of noise in x-ray images is quantum noise. It fol-lows a Poisson process that has a standard deviation of N , where N is the number of photons that arrive at the detector pixel. us, the resulting signal-to-noise ratio is ( )N N N/ = , that is, the more photons arrive at the detector, the higher the signal-to-noise ratio. As the number
CONTENTS 8.1 Introduction 113 8.2 Noise in X-Ray Images 114
8.2.1Noise in Projection Data 114 8.2.2Noise in Reconstructed Image Data 115
8.3Denoising Methods 118 8.3.1Gaussian Filter 118 8.3.2Bilateral Filter 120 8.3.3Joint Bilateral Filter 121 8.3.4Guided Filter 122 8.3.5Structure Tensor 123
8.4Denoising for CT 124 8.4.1Projection Domain 125 8.4.2Iterative Reconstruction 125 8.4.3Reconstruction Domain 125
8.5 Summary 126 References126
of measured photons is proportional to the number of emitted photons at the source, image noise is inversely proportional to the dose that is applied to the patient. us, methods for the reduction of image noise also enable a reduction in patient dose. For this reason, noise reduction techniques have been investigated in the literature extensively.
