A post-reconstruction method to correct cupping artifacts in cone beam breast computed tomography

Area of Science:

• Medical Imaging
• Radiology
• Image Processing

Background:

• Cone beam breast CT (CBCT) images suffer from cupping artifacts caused by scattered radiation.
• These artifacts manifest as background nonuniformities, hindering grayscale windowing and volume segmentation.
• Existing methods struggle to adequately address these specific artifacts in breast imaging.

Purpose of the Study:

• To develop and validate a specialized method for correcting background nonuniformity in CBCT breast images.
• To mitigate the impact of cupping artifacts on image analysis and visualization tasks.

Main Methods:

• Modeled cupping artifacts as a circularly symmetric additive background signal profile.
• Estimated radial background variations using adipose tissue signal measurements in front-view images.
• Implemented automated signal sampling in polar coordinates and background trend fitting for correction value calculation.

Main Results:

• The developed correction technique significantly improved signal uniformity in both front and side-view CBCT slices.
• Demonstrated a notable reduction in intraslice and interslice variations of adipose tissue CT numbers.
• Successfully addressed cupping artifacts, enhancing image quality for diagnostic purposes.

Conclusions:

• The proposed background nonuniformity correction method effectively reduces cupping artifacts in cone beam breast CT.
• This technique offers improved image quality, supporting more reliable quantitative analysis and visualization of breast tissue.
• The method shows promise for enhancing the clinical utility of CBCT in breast imaging applications.