The improved image enhancement method for complex shear modulus (CSM) proposed in this study combines two successive filtering operations: a median filter in the spatial domain and a band-pass filter (BPF) in the frequency domain. First, the finite-difference time-domain (FDTD) approach is used to simulate the propagation of shear waves. A BPF is used to minimize both low- and high-frequency noise by preserving signals centered around the excitation frequency and removing unwanted frequency components. The Algebraic Helmholtz Inversion (AHI) algorithm is then used to rebuild the image. A median filter is used as a post-processing step to further minimize spatial distortions and improve the clarity of anatomical boundaries. As demonstrated by lower normalized error metrics and better tissue contrast, numerical data show that the combined filtering scheme significantly increases the quantitative accuracy of elasticity and viscosity estimates. Thus, the suggested approach offers a strong and practical paradigm for improving the performance of real-time CSM imaging in clinical settings.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

A Two-Filter Approach for Enhanced Shear Modulus Imaging: Combining Frequency and Spatial Domain Filtering

  • Thi-Theu Luong,
  • Quang-Huy Tran,
  • Duc-Tan Tran,
  • Nguyen Ngoc Linh

摘要

The improved image enhancement method for complex shear modulus (CSM) proposed in this study combines two successive filtering operations: a median filter in the spatial domain and a band-pass filter (BPF) in the frequency domain. First, the finite-difference time-domain (FDTD) approach is used to simulate the propagation of shear waves. A BPF is used to minimize both low- and high-frequency noise by preserving signals centered around the excitation frequency and removing unwanted frequency components. The Algebraic Helmholtz Inversion (AHI) algorithm is then used to rebuild the image. A median filter is used as a post-processing step to further minimize spatial distortions and improve the clarity of anatomical boundaries. As demonstrated by lower normalized error metrics and better tissue contrast, numerical data show that the combined filtering scheme significantly increases the quantitative accuracy of elasticity and viscosity estimates. Thus, the suggested approach offers a strong and practical paradigm for improving the performance of real-time CSM imaging in clinical settings.