A Shearlet-based Algorithm for Quantum Noise Removal in Low-dose
CT Images
Aguan Zhang
a
, Huiqin Jiang
*a
, Ling Ma
b
, Yumin Liu
a
, Xiaopeng Yang
c
a
School of Information Engineering and Digital Medical Image Technique Research Center,
Zhengzhou University, Zhengzhou, China;
b
Fast Corporation 2791-5 Shimoturuma Yamoto,
Kanagawa, Japan;
c
Department of Equipment of the First Affiliated Hospital of Zhengzhou
University and Zhengzhou Engineering Technology Research Center for Medical Informatization,
Zhengzhou, China
ABSTRACT
Low-dose CT (LDCT) scanning is a potential way to reduce the radiation exposure of X-ray in the population. It is
necessary to improve the quality of low-dose CT images. In this paper, we propose an effective algorithm for quantum
noise removal in LDCT images using shearlet transform. Because the quantum noise can be simulated by Poisson
process, we first transform the quantum noise by using anscombe variance stabilizing transform (VST), producing an
approximately Gaussian noise with unitary variance. Second, the non-noise shearlet coefficients are obtained by adaptive
hard-threshold processing in shearlet domain. Third, we reconstruct the de-noised image using the inverse shearlet
transform. Finally, an anscombe inverse transform is applied to the de-noised image, which can produce the improved
image. The main contribution is to combine the anscombe VST with the shearlet transform. By this way, edge
coefficients and noise coefficients can be separated from high frequency sub-bands effectively. A number of experiments
are performed over some LDCT images by using the proposed method. Both quantitative and visual results show that the
proposed method can effectively reduce the quantum noise while enhancing the subtle details. It has certain value in
clinical application.
Keywords: low-dose CT images, quantum noise, Poisson noise, anscombe transform, shearlet transform
1. INTRODUCTION
Computed Tomography (CT) has made a great contribution to clinical diagnosis. X-rays is harmful to human health.
Minimizing radiation dose as more as possible has been a significant concern in CT imaging filed. Low-dose CT
(LDCT) scanning is a potential way to reduce X-ray radiation dose. However, it will produce strong quantum noise
which can affect the CT image quality and diagnostic accuracy seriously
1
. Therefore, it is necessary to remove the
quantum noise in LDCT images.
One scheme for removing the quantum noise is post-processing. Since the quantum noise is related to the value of CT
images, conventional
spatial domain denoising methods are not effective. Multi-scale transforms methods such as the
wavelet, the curvelet and the shearlet transforms, which can expand images into multi-scale and multi-direction, have
better
denoising performance than spatial domain methods. Besides, the quantum noise can be simulated by Poisson
processes
2.
Therefore, our research purpose is to remove the Poisson noise using multi-scale transform method for
improving LDCT images quality.
Wavelet transform is one of popular multi-scale transform methods for medical images denoising
3
. However, most of
developing wavelet-based denoising algorithms assume the noise following Gaussian distribution and estimate the local
statistics of image pixels, which are not effective for LDCT images. Recent years, there are some wavelet-based methods
have been proposed for removing the Poisson noise
4-5
. The main idea is first to transform Poisson noise to near Gaussian
noise using Variance stabilization transform (VST). Second, the noise is removed by using a conventional Gaussian noise
denoising algorithm. Finally, the de-noised image is obtained by using an inverse VST. For example, B.Zhang et al.
proposed a method combining VST with Wavelet, Ridgelet and Curvelet, which is efficient and sensitive in detecting
*
Email iehqjiang@zzu.edu.cn; Phone 86-371-67739503
Elsa D. Angelini, Proc. of SPIE Vol. 9784, 97843O · © 2016 SPIE
Proc. of SPIE Vol. 9784 97843O-1