Breast density quantification with cone-beam CT: a post-mortem study.

Area of Science:

• Medical Imaging
• Radiology
• Biomedical Engineering

Background:

• Accurate breast density quantification is crucial for breast cancer risk assessment and mammography screening.
• Existing imaging techniques face challenges in precise and reproducible breast density measurement.
• Chemical analysis provides a gold standard for determining tissue composition.

Purpose of the Study:

• To evaluate the feasibility of quantifying breast density using cone-beam CT (CBCT).
• To compare the performance of fuzzy c-means (FCM) clustering against histogram thresholding for breast density quantification.
• To validate image-based segmentation methods against gold standard chemical analysis of breast tissue composition.

Main Methods:

• Forty post-mortem breasts were imaged using a flat-panel based CBCT system at 50 kVp.
• Breast density was quantified using standard histogram thresholding and an automatic FCM segmentation algorithm.
• Breasts were chemically decomposed into water, lipid, and protein post-imaging for gold standard % fibroglandular volume (%FGV) determination.

Main Results:

• Both FCM and histogram thresholding demonstrated high precision in breast density quantification.
• FCM achieved a higher correlation (Pearson's r = 0.983) with chemical analysis %FGV compared to histogram thresholding (r = 0.968).
• FCM reduced the standard error of the estimate from 3.92% to 2.45% and decreased inter-observer variation.

Conclusions:

• CBCT imaging combined with FCM segmentation is a highly accurate and precise method for breast density quantification.
• Chemical analysis serves as a reliable gold standard for validating breast density measurement techniques.
• The FCM algorithm offers improved efficiency and reduced variability over conventional histogram thresholding for breast density assessment.