Affiliations: School of Geosciences and Info-Physics, Central South
University, Changsha, Hunan, China | Medical Imaging Center, Shaanxi Provincial People's
Hospital, Xi'an, Shaanxi, China
Note: [] Corresponding author: Gang Yu, School of Geosciences and
Info-Physics, Central South University, Changsha 41003, Hunan, China. Tel.: +86
13077327285; E-mail: yugang.2000@163.com
Abstract: PURPOSE: To develop a novel scatter correction method without
additional patient dose for dual-energy digital mammography (DEDM) to reduce
scatter's impacts and enhance microcalcification detectability in dual-energy
X-ray subtraction image. METHODS: Combining scatter radiation
is lower spatial frequency component and calcifications are sparsely
distributed in digital mammogram, we develop a new scatter correction strategy.
First, an adaptive sampling scheme is presented to find possible
noncalcification (zero calcification) pixels. Then the maximum likelihood
expectation maximization (MLEM) algorithm is applied to evaluate initial
scatter surface. The accurate scatter radiation of sampling pixels is obtained
by solving dual-energy computational formula with zero calcification constraint
and scatter surface constraint. RESULTS: After scatter
correction, the scatter-to-primary ratio (SPR) of wedge phantom is reduced from
∼36.0% to ∼3.1% for low-energy (LE) image and ∼29.6%
to ∼0.6% for high-energy (HE) image. For step phantom, the SPR is
reduced from ∼42.1% and ∼30.3% to ∼3.9% and ∼0.9% for LE and HE image,
respectively. The calcification contrast-to-noise
ratio is improved by two orders of magnitudes in calcification images. CONCLUSIONS: The proposed method shows an excellent performance
on scatter reduction and calcification detection. Compared with hardware based
scatter correction strategy, our method need no extra exposure and is easy to implementation.
Keywords: Scatter correction, dual-energy mammography, calcification, monte carlo
