Backscatter coefficient estimation using tapers with gaps

Area of Science:

• Quantitative ultrasound imaging
• Biomedical acoustics
• Signal processing

Background:

• Accurate estimation of quantitative ultrasound parameters like effective scatterer diameter (ESD) and effective acoustic concentration (EAC) relies on the assumption of diffuse scatterers.
• Non-diffuse scatterers introduce bias and variance in backscatter coefficient (BSC) estimations, hindering disease classification performance.

Purpose of the Study:

• To develop a method for mitigating the impact of non-diffuse echoes in quantitative ultrasound.
• To preserve diffuse signal components for accurate BSC parameter estimation.
• To evaluate the trade-offs between specular scatterer mitigation and introduced bias/variance.

Main Methods:

• A novel windowing technique was employed to identify and isolate non-diffuse backscattered signal sections.
• The method provides BSC estimates exclusively from diffuse echo signals.
• Experiments were conducted using physical phantoms to validate the proposed BSC estimation techniques.

Main Results:

• Discrete prolate spheroidal (PR) tapers with gaps demonstrated superior performance in minimizing BSC error compared to Hanning tapers.
• PR tapers with six gaps achieved a mean square error of approximately 0.2, significantly lower than Hanning tapers (approx. 1.0).
• Specific PR taper configurations (Nω = 1 and Nω = 4) minimized amplitude and shape bias, respectively.

Conclusions:

• The developed method effectively mitigates non-diffuse scatterer effects in quantitative ultrasound.
• PR tapers are recommended for optimizing BSC estimation accuracy, with different configurations suited for ESD and EAC.
• This approach enhances the reliability of ultrasound-based tissue characterization.