A Modified Higher-Order Singular Value Decomposition Framework With Adaptive Multilinear Tensor Rank Approximation for Three-Dimensional Magnetic Resonance Rician Noise Removal.

Li Wang,Wen S Hou,Di Xiao,Lin Chen,Xiao Y Wu

doi:10.3389/fonc.2020.01640

Abstract

The magnetic resonance (MR) images are acknowledged to be inevitably corrupted by Rician distributed noise, which adversely affected the image quality for diagnosis purpose. However, the traditional denoising methods may recover the images from corruptions with severe loss of detailed structure and edge information, which would affect the lesion detections and diagnostic decision making. In this study, we challenged improving the Rician noise removal from three-dimensional (3D) MR volumetric data through a modified higher-order singular value decomposition (MHOSVD) method. The proposed framework of MHOSVD involved a parameterized logarithmic nonconvex penalty function for low-rank tensor approximation (LRTA) algorithm optimization to suppress the image noise in MR dataset. Reference cubes were extracted from the noisy image volume, and block matching was performed according to nonlocal similarity for a fourth-order tensor construction. Then the LRTA problem was implemented by tensor factorization approaches, and the ranks of unfolding matrices along different modes of the tensor were estimated utilizing an adaptive nonconvex low-rank method. The denoised MR images were finally restored through aggregating all recovered cubes. We investigated the proposed algorithm MHOSVD on both the synthetic and real clinic 3D MR images for Rician noise removal, and relative results demonstrated that the MHOSVD can recover images with fine structures and detailed edge preservation with heavy noise even as high as 15% of the maximum intensity. The experimental results were also compared along with several classical denoising methods; the MHOSVD exhibited a sufficient improvement in noise-removal performance at various noise conditions in terms of different measurement indices such as peak signal-to-noise ratio and structural similarity index metrics. Based upon the comparison, the proposed MHOSVD has proved a relative state-of-the-art performance with excellent detailed structure reservation.

Highlights

Magnetic resonance imaging (MRI), as a widely accepted noninvasive imaging modality, was acknowledged as a useful diagnostic tool with high resolution and excellent contrast sensitivity to anatomical properties [1]
Proposed Model describes our algorithm proposed for Rician noise removal based upon adaptive HOSVD framework
We validated the performance of the proposed algorithm on both the synthetic and clinical 3D magnetic resonance (MR) images

Summary

INTRODUCTION

Magnetic resonance imaging (MRI), as a widely accepted noninvasive imaging modality, was acknowledged as a useful diagnostic tool with high resolution and excellent contrast sensitivity to anatomical properties [1]. An effective denoising algorithm specific for Rician noise based upon three-dimensional (3D) MR image datasets can play a fundamental role to improve the diagnostic accuracy for both the radiologists and the CAD tools. Filter-based approaches constructed the noise-reduction schemes with linear or nonlinear filters, such as anisotropic diffusion and total variation (TV) techniques, to improve image quality with edge preservation considerations [5,6,7,8]. The well-known NLM filter that Buades et al [24] proposed was one classical method that considered exploiting nonlocal patch similarity for noise removal [24]. Experiments were performed based on various MR images, obtained from either computer synthesis or clinical screening, to investigate denoising improvements of proposed framework against several advanced algorithms with the following novelties: 1. Last but not the least, we draw our conclusion in Conclusions

BACKGROUND

EXPERIMENTS AND RESULTS

CONCLUSIONS