Comparing measurements of breast density

Area of Science:

• Radiology
• Medical Imaging
• Breast Imaging

Background:

• Breast density assessment is crucial for mammography interpretation and breast cancer risk prediction.
• Area-based (e.g., Six-Category Classification, SCC) and volumetric-based (e.g., Standard Mammogram Form, SMF) methods are used for breast density estimation.
• Previous work showed general agreement between SCC and SMF, but significant discrepancies occur in some cases.

Purpose of the Study:

• To investigate the reasons for significant differences in breast density estimates between area-based (SCC) and volumetric-based (SMF) mammographic methods.
• To analyze the contribution of technical errors versus fundamental methodological differences to these discrepancies.

Main Methods:

• Analysis of mammograms from 657 patients comparing SCC and SMF assessments.
• Identification of cases with substantial differences between SCC and SMF estimates.
• Evaluation of potential error sources including segmentation algorithms, human categorization, image quality, and inherent method differences.

Main Results:

• Approximately 12% of patients exhibited large differences between SCC and SMF breast density assessments.
• Errors in SMF segmentation, SCC categorization, or image exposure accounted for 2.7% of discrepancies.
• Fundamental differences between area- and volume-based techniques were responsible for 9.3% of the observed large differences.

Conclusions:

• Significant discrepancies between SCC and SMF breast density estimates arise from both technical errors and inherent differences in the underlying physical models.
• Introducing breast thickness categorization for SMF estimates may reduce discrepancies by half, but a residual 6% difference remains.
• The assumption that volumetric measures (SMF) possess similar breast cancer risk prediction capabilities to area-based measures (SCC) requires further investigation.