3-D Perfusion Imaging Using Principal Curvature Detection Rendering.

Area of Science:

• Medical imaging
• Ultrasound technology
• Oncologic diagnostics

Background:

• Three-dimensional contrast-enhanced ultrasound (3-D CEUS) offers superior tumor lesion detection over 2-D methods.
• Current 3-D CEUS limitations include slow imaging and undersampling, hindering visualization of crucial tumor feeding vessels.
• Volumetric Doppler angiography faces real-time challenges due to Doppler ensemble length constraints.

Purpose of the Study:

• To demonstrate the feasibility of real-time volumetric CEUS angiography under imaging constraints.
• To improve the visualization of tumor vascularity in 3-D CEUS.
• To overcome limitations in resolving feeding vessels and background perfusion.

Main Methods:

• Utilized plane-wave imaging to accelerate volumetric imaging frame rates.
• Applied principal curvature detection to enhance 3-D rendering of noisy ultrasound angiograms.
• Employed singular value decomposition for identifying quasi-stationary capillary perfusion.

Main Results:

• Principal curvature detection significantly improved 3-D rendering of ultrasound angiograms without spatial resolution loss.
• The method allowed for effective visualization with a reasonable Doppler ensemble size.
• Singular value decomposition successfully identified capillary perfusion patterns.

Conclusions:

• Volumetric CEUS angiography is feasible under real-time imaging constraints.
• Principal curvature detection is a valuable tool for enhancing 3-D ultrasound angiograms in oncology.
• The developed methods improve the characterization of tumor vascularity using 3-D CEUS.