Fast 3D Ultrasound Localization Microscopy via Projection-based Processing Framework.

Area of Science:

• Medical Imaging
• Ultrasound Technology
• Biomedical Engineering

Background:

• Three-dimensional ultrasound localization microscopy (3D ULM) offers detailed vascular visualization for improved diagnostics.
• Clinical adoption of 3D ULM is hindered by extensive computational requirements and long processing times for full 3D reconstructions.
• Current methods face challenges with large voxel counts, limiting real-time application and increasing operator dependence.

Purpose of the Study:

• To develop a computationally efficient 3D ULM framework for in vivo imaging.
• To reduce the processing demands of 3D ULM while maintaining diagnostic accuracy.
• To explore the potential for real-time feedback in 3D ULM-guided procedures.

Main Methods:

• A row-column array (RCA)-based 3D ULM pipeline was reformulated using efficient 2D operations for beamforming, clutter filtering, motion estimation, and microbubble localization.
• The proposed method processes each 0.75-s ensemble at a 400 Hz volume rate, covering a 25×27.4×27.4 mm³ volume.
• Reconstruction was performed on a single NVIDIA RTX A6000 Ada GPU.

Main Results:

• The framework achieved reconstruction in 0.52 seconds (70% of acquisition time), significantly faster than conventional 3D processing.
• ULM image quality comparable to conventional 3D methods was maintained.
• Quantitative analysis showed a structural similarity index (SSIM) of 0.93 and voxel-wise velocity agreement (slope=0.93, R²=0.88).

Conclusions:

• The developed computational framework enables efficient 3D ULM processing, overcoming previous limitations.
• This approach maintains high image quality and quantitative accuracy, comparable to traditional methods.
• The study demonstrates the potential for real-time feedback in 3D ULM, enhancing scanning robustness and accelerating clinical workflows.