Robust Estimation of Displacement in Real-Time Freehand Ultrasound Strain Imaging

Area of Science:

• Medical Imaging
• Biomedical Engineering
• Ultrasound Technology

Background:

• Real-time strain imaging in ultrasound elastography is crucial for diagnosing tissue stiffness.
• Existing methods for displacement estimation often struggle with robustness and speed, limiting clinical application.
• Accurate displacement estimation is fundamental for generating reliable strain images.

Purpose of the Study:

• To develop a novel and efficient approach for robust displacement estimation in real-time ultrasound elastography.
• To enhance the quality and reliability of strain images generated from freehand ultrasound data.
• To improve upon existing methods by increasing robustness without compromising speed.

Main Methods:

• Utilizing pre- and post-deformation ultrasound images for displacement estimation.
• Defining a quality factor for image lines to identify a seed line for displacement mapping.
• Developing an analytical framework for coarse-to-fine displacement estimation with subsample precision.
• Implementing a fast strategy for seed line displacement estimation and propagation.
• Dynamically identifying a new seed line when quality falls below a threshold to maintain robustness.

Main Results:

• Achieved high signal-to-noise-ratio and contrast-to-noise-ratio values across a wide range of average strain levels (1%-10%).
• Demonstrated robustness against corrupt and decorrelated data in phantom experiments.
• Produced high-quality strain images at a rate of 20 frames/s on conventional CPUs, outperforming existing real-time methods.
• Successfully generated reliable displacement maps and strain images using the novel seed line approach.

Conclusions:

• The presented method offers a robust and efficient solution for real-time displacement estimation in ultrasound elastography.
• The quality factor-based seed line selection and dynamic updating enhance image quality and reliability.
• This approach is superior to existing real-time methods, enabling high-quality strain imaging even at higher strain levels and faster frame rates.