Revisiting artifact spread function evaluation in digital breast tomosynthesis systems

Area of Science:

• Medical Imaging
• Radiology
• Biomedical Engineering

Background:

• Digital Breast Tomosynthesis (DBT) is crucial for breast cancer detection.
• Zipper artifacts in DBT images complicate accurate diagnosis.
• A standardized methodology for evaluating zipper artifacts is lacking.

Purpose of the Study:

• To identify the optimal method for characterizing zipper artifacts.
• To use Full Width at Half Maximum (FWHM) of the Artifact Spread Function (ASF) as a metric.
• To employ aluminum sphere phantoms for generating zipper artifacts.

Main Methods:

• Compared zipper artifact characterization on GE Senographe Pristine (SP-GE) and Hologic Selenia Dimensions (SD-HO) systems.
• Calculated ASF using maximum value, mean value, and a hybrid method.
• Evaluated ASF generated by aluminum, lead, tungsten, and ceramic markers, and a commercial phantom.

Main Results:

• The hybrid ASF method was most reliable for the SP-GE system.
• The mean value ASF method best suited the SD-HO system.
• Ceramic and lead markers better replicated clinical zipper artifacts across most methods.

Conclusions:

• Selecting an appropriate artifact evaluation methodology depends on the specific DBT system.
• Standardized approaches are needed to account for system-specific differences in artifact assessment.
• Accurate artifact assessment is vital for reliable breast cancer diagnosis using DBT.