Enhancing ultrasonic attenuation images through multi-frequency coupling with total nuclear variation

Area of Science:

• Medical Imaging
• Biomedical Engineering
• Acoustics

Background:

• Quantitative ultrasound (QUS) is a non-invasive imaging technique for medical diagnosis, utilizing acoustical parameters like attenuation coefficient slope.
• Existing methods, such as total variation spectral log difference (TVSLD), denoise spectral log ratios but lack inter-channel coupling.

Purpose of the Study:

• To propose a multi-frequency joint framework to enhance the quality of attenuation images in quantitative ultrasound.
• To introduce and evaluate a modification based on total nuclear variation (TNV) for improved QUS imaging.

Main Methods:

• Developed a multi-frequency joint framework that couples information across frequency channels by exploiting structural similarities in spectral ratios.
• Introduced a Total Nuclear Variation (TNV) modification to the framework.
• Compared TNV performance against the TVSLD method using simulated phantoms, experimental phantoms, and in vivo breast tissue samples.

Main Results:

• The TNV method demonstrated superior performance, producing enhanced attenuation coefficient slope maps with reduced artifacts and stable error.
• TNV achieved significant contrast-to-noise ratio (CNR) improvements: 34% in simulations, 38% in experimental phantoms, and 89% in vivo.
• Results indicate TNV's potential to enhance visual clarity and depiction in QUS attenuation images.

Conclusions:

• The proposed Total Nuclear Variation (TNV) multi-frequency joint framework significantly enhances quantitative ultrasound imaging quality.
• TNV offers superior artifact reduction and CNR improvement compared to TVSLD, particularly in in vivo applications.
• This method shows promise for improving diagnostic accuracy through clearer attenuation coefficient slope mapping.