Abstract:Purpose: Previously, the authors developed a dual-energy (DE) acquisition technique for a photon-counting digital breast tomosynthesis (DBT) imaging system. Low-energy (LE) and high-energy (HE) images are acquired in a single scan by covering alternate slits of a multislit prepatient collimator with Sn and Cu, respectively. A theoretical model was used to optimize the technique. In this article, an experimental validation of this technique is presented. Methods: Experiments were performed on a prototype DBT system. LE and HE projection images were acquired sequentially; either a Sn or a Cu filter was positioned in the filter holder at the exit window of the x-ray tube. Sn filters from 0.113 to 0.242 mm thick and Cu filters from 0.103 to 0.267 mm were used. The images were acquired with a W target at 49 kV. Tomographic images, hereafter referred to as DBT images, were reconstructed using a shift-and-add algorithm. DE-DBT images were obtained by weighted logarithmic subtraction of the LE and HE images. Weighting factors w(t) that optimally cancel breast tissues with two different glandularities were assessed for 20-80 mm thick phantoms with 0%, 50%, and 100% glandularity. The mean and standard deviation in the per-pixel signal intensity (SI) were calculated in the DBT images. These data were used to calculate signal-difference-to-noise ratios (SDNRs) between iodine enhanced and nonenhanced polymethyl methacrylate backgrounds. To illustrate the feasibility of the technique, DE-DBT images of a structured phantom containing iodine disks were assessed. The experimental results were compared against the values obtained from a theoretical model of the imaging system. Results: The average difference between theoretical and experimental w(t) was found to range from 8% to 21%. Experimental w(t) values increase with phantom thickness and Cu thickness, depend somewhat on Sn thickness, and vary more as a function of breast composition in thick breasts than in thin breasts. Theoretical and experimental mean and standard deviation in the per-pixel SI differ by -7% to 10% and by -3% to 4%. Theoretical and experimental SDNR values differ, on average, by 1.5%. Iodine concentrations can be predicted from SDNR; the relationship can be accurately fit to a quadratic. In the images of the structured phantom, iodine concentrations of 1 mg/cm2 and larger are discernable. Conclusions: The strong agreement between experimental and theoretical results in this article indicates that the authors' computer model is accurate.

Preliminary Attempt on Maximum Likelihood Tomosynthesis Reconstruction of DEI Data

SU‐FF‐I‐39: A Computer Simulation Platform for the Optimization of Breast Tomosynthesis System

An ART Iterative Reconstruction Algorithm for Computed Tomography of Diffraction Enhanced Imaging

Computed Tomography Algorithm Based on Diffraction-Enhanced Imaging Setup

Enhanced Imaging Of Microcalcifications In Digital Breast Tomosynthesis Through Improved Image-Reconstruction Algorithms

Low-Dose Digital Tomosynthesis Reconstruction Based On Field Emission Flat-Panel X-Ray Source Array

A Preliminary Study Of Image Reconstruction From Low-Dose Data In Dedicated Breast Ct

Optimization methods for very accurate Digital Breast Tomosynthesis image reconstruction

An Adaptive Optimum Contrast CT Image Acquisition for Improved Spectrum Estimation-Guided DECT reconstruction

A Novel Digital Tomosynthesis (DTS) Reconstruction Method Using a Deformation Field Map

Deep Learning Framework for Digital Breast Tomosynthesis Reconstruction

Dei-Based Phase-Contrast Tomosynthetic Experiment on Biological Samples with High Resolution X-Ray Ccd Camera

Optimization of a dual-energy contrast-enhanced technique for a photon-counting digital breast tomosynthesis system: II. An experimental validation

MISD-IR: material-image subspace decomposition-based iterative reconstruction with spectrum estimation for dual-energy computed tomography

Performance evaluation of deep learning image reconstruction algorithm for dual-energy spectral CT imaging: A phantom study

Preliminary Study of Image Reconstruction for Breast DECT with Limited-Angular-Range Data

Material decomposition for simulated dual-energy breast computed tomography via hybrid optimization method

Towards dose reduction for dual-energy CT: A non-local image improvement method and its application

Differential Phase-Contrast Tomosynthetic Experimental System with Weakly Coherent Hard X-rays

SU-E-I-149: Digital Breast Tomosynthesis Using 2D Source Scanning Patterns: A Simulation Study

Tomosynthesis reconstruction from multi-beam X-ray sources