Microcalcification detection using cone-beam CT mammography with a flat-panel imager

Area of Science:

• Medical Imaging
• Radiology
• Biophysics

Background:

• Microcalcifications are key indicators of early breast cancer.
• Current mammography techniques have limitations in microcalcification detection.
• Advancements in imaging technology are needed to improve diagnostic accuracy.

Purpose of the Study:

• To evaluate microcalcification detectability using computed tomography (CT) mammography with a flat-panel imager.
• To determine the impact of detector pixel size on microcalcification detection.
• To assess the feasibility of CT mammography for screening.

Main Methods:

• A computer simulation modeled an amorphous-silicon, CsI based flat-panel imager system.
• The breast was simulated as a hemi-ellipsoid with mixed adipose and glandular tissue.
• Microcalcifications were modeled as calcium carbonate spheres.
• A receiver operating characteristic (ROC) study was performed by physicist observers.

Main Results:

• CT mammography with a flat-panel detector can detect most microcalcifications at screening mammography dose levels.
• Microcalcifications ≥ 0.175 mm diameter were detected with an average area under the ROC curve (AUC) > 0.95 using 0.1 or 0.2 mm pixelized detectors.
• An optimal detector pixel size of approximately 0.2 mm was identified for microcalcification detection.

Conclusions:

• CT mammography utilizing flat-panel detectors shows significant potential for detecting microcalcifications.
• Detector pixel size is a critical factor influencing microcalcification detectability and overall imaging efficiency.
• Further research and clinical validation are warranted to establish CT mammography as a viable screening tool.