Motion estimation-based image enhancement in ultrasound imaging

Area of Science:

• Medical Imaging
• Ultrasound Technology
• Image Processing

Background:

• High-resolution medical ultrasound (US) imaging is crucial for diagnosis but faces challenges.
• Existing super-resolution (SR) techniques from optical imaging are not directly applicable to US due to tissue interaction and speckle.
• Limited research exists on SR image reconstruction specifically for US imaging.

Purpose of the Study:

• To overcome the limitations of standard SR methods in ultrasound imaging.
• To improve the resolution and quality of medical ultrasound images through advanced post-processing.
• To adapt and refine multiframe SR techniques for US applications.

Main Methods:

• Investigated and refined motion estimation methods specifically adapted for ultrasound imaging.
• Improved the registration step within common multiframe super-resolution techniques.
• Evaluated the proposed technique on simulated US images, phantom data, and in-vivo thyroid sequences.

Main Results:

• Achieved resolution gains of 1.41 on phantom sequences and 1.12 on in-vivo thyroid sequences.
• Demonstrated significant quantitative and qualitative improvements compared to classical SR methods.
• Increased contrast-to-noise ratio by 27% for simulated and 13% for experimental US images.

Conclusions:

• The refined SR technique effectively enhances spatial resolution in medical ultrasound imaging.
• The proposed method offers a viable solution for improving US image quality beyond current limitations.
• This advancement holds potential for improved diagnostic accuracy in various US applications.