Gauging precision of breast density measurement in spectral mammography

Dual energy material composition is twice as precise as other interpretation techniques when it comes to breast density assessment, according to results of a study published online May 29 in Academic Radiology.

The importance of mammographic breast density as an indicator of a patient’s future risk of breast cancer has spurred efforts within the medical community to explore new imaging techniques in search of more accurate methods of assessing breast tissue density, according to lead author Sabee Molloi, PhD, and his colleagues at the University of California, Irvine.

“The importance of quantitative breast density assessment has been highlighted by a previous report indicating that for every 1 percent increase of mammographic breast density, there is a 2 percent increase of the relative risk for breast cancer,” wrote Molloi and colleagues. “Therefore, improved methods of measuring breast density could potentially be helpful in more accurately quantifying breast cancer risk and monitoring changes in risk over time.”

Molloi and his team set out to compare how precise various interpretation techniques were in measuring mammographic breast density. To do so, they performed a retrospective study on the spectral mammography images from 92 screening patients between the ages of 50–69. Breast density was estimated using four distinct techniques: a 10-radiologist reader assessment, standard histogram thresholding, fuzzy C-mean algorithm and dual-energy material decomposition. Breast density correlation between left and right breasts was used to determine the precision of breast composition assessment.

With dual-energy material decomposition, the physical differences in mass attenuation coefficients of glandular and adipose tissues as a function of beam energy are analyzed and the thickness of tissue is quantified on a pixel-by-pixel basis, explained the authors. 

Their results showed that dual-energy material composition showed the highest correlation, with relative standard error of estimate for breast density measurements of 1.95 for radiologist rankings, 2.87 for standard histogram thresholding (Cumulus), 2.07 for fuzzy C-mean algorithm,1.00 or dual-energy material decomposition.

“The results indicate that the precision of dual energy material decomposition was approximately factor of two higher than the other techniques with regard to better correlation of breast density measurements from right and left breasts,” the authors concluded. “Improved quantification of breast density is expected to further enhance its utility as a risk factor for breast cancer.”