Ultrasound speckle reduction using modified Gabor filters

Area of Science:

• Medical Imaging
• Signal Processing
• Biomedical Engineering

Background:

• Speckle artifact in B-mode ultrasound images hinders interpretation.
• Traditional split spectrum processing (SSP) for speckle reduction causes significant resolution loss due to narrow-band filters.

Purpose of the Study:

• To introduce a novel method for speckle reduction in ultrasound medical imaging.
• To overcome the resolution loss associated with existing speckle reduction techniques.

Main Methods:

• Developed a novel directive filtering (DF) method using wideband 2-D directive filters based on modified Gabor functions.
• Applied filters to 2-D radio-frequency (RF) data to create directionally filtered B-mode images.
• Compounded filter outputs to produce a final image with reduced speckle and enhanced structure.

Main Results:

• The proposed directive filtering (DF) method avoids resolution loss by using wideband filters with the same bandwidth as the original image.
• Achieved an average signal-to-noise ratio (SNR) enhancement of 2.26 times with simulated data.
• Achieved an average SNR enhancement of 1.18 times with real clinical data.

Conclusions:

• The directive filtering (DF) method effectively reduces speckle artifact in ultrasound images.
• This technique enhances image structure and clarity without compromising image resolution.
• The proposed method shows promise for improving the diagnostic quality of ultrasound imaging.