Multicycle Temporal Compounding Sonography for Despeckling at Specified Respiratory Phases

Area of Science:

• Medical Imaging
• Ultrasound Technology
• Image Processing

Background:

• Abdominal ultrasound imaging is often hindered by speckle noise and motion artifacts, particularly in intensive care units (ICUs).
• Current methods frequently require patients to hold their breath, which can be challenging or impossible in critical care settings.
• Motion artifacts and speckle noise degrade image quality, impacting diagnostic accuracy.

Purpose of the Study:

• To develop a breath-hold-free abdominal ultrasound method for improved image quality in clinical and ICU settings.
• To address and mitigate speckle noise and motion artifacts in ultrasound images.
• To enhance the reliability and efficiency of abdominal ultrasound examinations.

Main Methods:

• A novel respiratory phase-guided multicycle temporal compounding technique was employed.
• The method identifies motion states across respiratory cycles to fuse images for effective despeckling.
• A synergistic framework combined anisotropic diffusion for speckle removal and guided filtering for structure transfer.

Main Results:

• The method achieved high performance metrics, including an average β of 0.931/0.930 and a Speckle Suppression and Mean Preservation Index (SMPI) of 0.483/0.584.
• Pratt's figure of merit was high at 0.842/0.827, indicating superior edge preservation and despeckling.
• The technique demonstrated computational efficiency comparable to established methods, with ~2.65s/iteration.

Conclusions:

• The developed breath-hold-free ultrasound method effectively suppresses speckle noise and motion artifacts.
• It outperforms existing despeckling methods in terms of edge preservation and overall image quality.
• This technique offers a streamlined workflow for abdominal ultrasound in critical care and routine practice without requiring breath-holding.