Breast density quantification in dual-energy mammography using virtual anthropomorphic phantoms

Area of Science:

• Medical Imaging
• Radiology
• Biophysics

Background:

• Breast density is a key breast cancer risk factor.
• Area-based density measurements lack accuracy; volumetric breast density (VBD) is preferred.
• Dual-energy mammography allows VBD calculation without shape assumptions.

Purpose of the Study:

• Evaluate a dual-energy decomposition technique for VBD measurement.
• Validate the technique using virtual anthropomorphic phantoms.
• Assess performance with clinical dual-energy mammography data.

Main Methods:

• Quantified VBD using dual-energy decomposition on simulated images of 150 virtual phantoms.
• Simulated phantoms with volumes from 50-709 mL and VBD from 15-60%.
• Validated a skin correction, assessed bias/variability, and compared to MRI in 14 clinical cases.

Main Results:

• Virtual phantom VBD measurements strongly correlated with nominal values.
• Skin correction reduced VBD bias to -2% and eliminated size variability.
• Clinical VBD agreement between dual-energy mammography and MRI was <10% (Spearman's rho = 0.68).

Conclusions:

• Material decomposition accurately measures VBD when skin is considered.
• Dual-energy mammography VBD shows agreement with MRI.
• VBD assessment via dual-energy imaging offers complementary information for mammography and tomosynthesis.