Multi-view 3D echocardiography compounding based on feature consistency.

Area of Science:

• Medical Imaging
• Cardiovascular Ultrasound
• Image Processing

Background:

• Echocardiography (echo) is crucial for cardiac imaging but faces limitations like artefacts, noise, and restricted field of view.
• Compounding multiple echo images can overcome these limitations by creating a higher-quality composite image.

Purpose of the Study:

• To develop and validate a novel compounding algorithm for echocardiography.
• To improve signal-to-noise ratio, contrast, and field of view while reducing artefacts in cardiac ultrasound images.

Main Methods:

• A new compounding algorithm was developed, weighting image information based on local feature coherence across overlapping multi-view 3D echocardiography images.
• The algorithm was validated using phantom, volunteer, and patient datasets, comparing results against original echo images and basic statistical methods (mean, maximum).

Main Results:

• The proposed method produced compounded images of quality comparable to direct acquisitions, even when original images were degraded by tissue artefacts.
• It achieved significant signal-to-noise and contrast improvements, approaching the performance of mean and maximum methods, respectively.
• Using ten images significantly improved quality over five, with clinicians preferring the compounded volumes for overall quality, field of view, and definition.

Conclusions:

• The developed compounding algorithm effectively enhances echocardiography image quality by addressing artefacts, noise, and field-of-view limitations.
• This method offers a valuable tool for improving cardiac ultrasound diagnostics and visualization.
• Clinical validation demonstrated a strong preference for the compounded images, highlighting their diagnostic utility.