Signal Detection of Point Targets Using Eigen-Images for Super-Resolution Ultrasound Imaging and Gas Vesicle...

Area of Science:

• Medical Imaging
• Biomedical Engineering
• Signal Processing

Background:

• Accurate detection of ultrasound contrast agents (microbubbles and gas vesicles) is crucial for super-resolution ultrasound imaging (SRUS) and precise localization.
• Clutter and noise in ultrasound data can hinder image quality and accurate signal detection.

Purpose of the Study:

• To develop and evaluate an automated eigen-image based signal detection method using singular value decomposition (SVD) and changepoint detection.
• To compare the eigen-image based method with existing techniques for microbubble (MB) signal selection and gas vesicle (GV) localization.

Main Methods:

• Singular value decomposition (SVD) and changepoint detection were employed to create an eigen-image based signal detection method.
• The method was validated using phantom and in vivo experiments, comparing MB signal detection and GV localization against established methods.

Main Results:

• The eigen-image based method demonstrated superior vessel density (VD) visualization and increased signal-to-noise ratio (SNR) for MB detection in both phantom and in vivo studies.
• It achieved more efficient localization of moving GVs compared to difference imaging, eliminating the need for landmark-based registration.

Conclusions:

• The eigen-image based method provides a reliable and automated approach for MB and GV signal detection in SRUS and point target localization.
• This technique enhances medical imaging by delivering high-quality vessel images and accurate localization of moving ultrasound contrast agents, with potential clinical and pre-clinical applications.