Fractional fourier-based filter for denoising elastograms

Area of Science:

• Medical Imaging
• Biomedical Engineering
• Signal Processing

Background:

• Ultrasound elastography estimates tissue properties from axial displacements.
• Calculating axial strains amplifies noise, hindering accurate analysis.
• Existing methods struggle with non-stationary displacement signals.

Purpose of the Study:

• To develop a novel denoising scheme for accurate axial strain estimation in ultrasound elastography.
• To address noise amplification issues inherent in displacement differentiation.
• To improve the contrast-to-noise ratio (CNR(e)) of elastograms.

Main Methods:

• A denoising method based on repeated filtering in fractional Fourier transform domains.
• Implementation of a time-varying cutoff threshold to handle signal non-stationarities.
• Utilizing a filter circuit with linear low-pass filters and fractional Fourier transforms.

Main Results:

• The proposed method effectively reduces noise in axial strain calculations.
• Achieved significant improvement in the contrast-to-noise ratio (CNR(e)) of elastograms.
• Outperformed conventional low-pass filtering techniques in denoising displacement signals.

Conclusions:

• The novel fractional Fourier transform-based denoising scheme enables accurate axial strain estimation.
• This technique enhances the diagnostic quality of ultrasound elastography images.
• The method offers a robust solution for analyzing non-stationary displacement signals in medical imaging.