Ultrasound speckle suppression and edge enhancement using multiscale nonlinear wavelet diffusion

Area of Science:

• Medical Imaging
• Signal Processing
• Image Analysis

Background:

• Ultrasound imaging is crucial for medical diagnosis but suffers from speckle noise, which degrades image quality.
• Existing speckle suppression methods often compromise edge preservation, hindering accurate interpretation.
• Edge enhancement is vital for delineating anatomical structures in ultrasound images.

Purpose of the Study:

• To introduce a novel multiscale nonlinear wavelet diffusion (MNWD) method for simultaneous ultrasound speckle suppression and edge enhancement.
• To adapt nonlinear diffusion principles within the dyadic wavelet transform framework for improved image processing.
• To evaluate the efficacy of MNWD in preserving structural details while reducing noise.

Main Methods:

• The proposed MNWD method utilizes the sparsity and multiresolution properties of wavelets.
• Iterative multiscale diffusion is applied to wavelet coefficients for speckle suppression.
• An iterative signal compensation process is employed for edge enhancement.

Main Results:

• The MNWD method demonstrated significant speckle suppression in both synthetic and real echocardiographic images.
• Edge enhancement was achieved concurrently with noise reduction, preserving important image features.
• Quantitative analysis showed superior performance over traditional denoising filters in noise suppression and structural preservation.

Conclusions:

• The MNWD method offers an effective solution for ultrasound speckle suppression and edge enhancement.
• The integration of wavelet transform and nonlinear diffusion provides a powerful framework for medical image processing.
• The method shows promise for improving the accuracy of subsequent image analysis tasks, such as segmentation.