Colin Auclair Dipl Eng, MSc, Presenter: Employee, General Electric Company

Jeffrey Shaw PhD, MS, Abstract Co-Author: Employee, General Electric Company

Mathias Cisaruk DIPLENG, Abstract Co-Author: Employee, General Electric Company

Remy Klausz DiplEng, Abstract Co-Author: Employee, General Electric Company

Henri Souchay PhD, Abstract Co-Author: Employee, General Electric Company

Conventional grids for mammography (MG) with septa perpendicular to the chest wall are not suitable for digital breast tomosynthesis (DBT) because the trajectory of the x-ray tube relative to the grid moves the source out of the grid focus line. The performance of a new 3D-grid, compatible with DBT and MG, was evaluated and its benefits compared with a conventional grid for MG, and with no grid at equal dose for DBT.

The septa of the 3D-grid are designed to be parallel to the tube sweep trajectory. To eliminate grid line visibility, the septa interspace is matched to the detector pixel pitch. During image acquisition, the 3D-grid moves with sub-mm amplitude, preserving tissue visibility at chestwall side. Grid performance was determined using methods derived from IEC 60627 standards, which allows to estimate the theoretical image quality improvement at same dose for average breasts. Evaluation of the image quality performance in 2D used a CDMAM phantom sandwiched between 2cm PMMA plates. For DBT, image quality performance was compared from the detectability of inserts in an ACR phantom imaged directly and on top of 2.5cm PMMA. All images were acquired on a Senographe Essential (GE Healthcare) using the same technique factors for each comparison.

For 5cm PMMA and at 28kV MoMo, primary transmission of the 3D-grid was 70.1%±0.5%, indicating that the grid contribution to SDNR² for average breast patient was 0.99±0.03. For MG, automatic CDMAM scoring (CDMAM Analyser, Artinis) provided similar IQF scores for the 3D-grid (122) and conventional grid (127) when using the same acquisition parameters. For DBT, at the central plane of the inserts, ACR phantom scores were similar with (13.3 ± 0.5) and without (13.2 ± 0.5) grid. When adding 2.5 cm PMMA, ACR phantom scores were higher with (10.3 ± 0.5) than without (8.3 ± 0.5) grid. In addition, the 3D-grid significantly improved signal uniformity throughout the phantom.  

The 3D-grid demonstrated a potential of improving detectability of features for breasts above the average thickness, while preserving the dose in DBT.

For breasts above the average thickness, most difficult to image in mammography, the 3D-grid offers scatter rejection benefits comparable to MG grids, yet capable of operating both in MG and in DBT.

Auclair, C, Shaw, J, Cisaruk, M, Klausz, R, Souchay, H, Introduction and Benefits of an Anti-scatter Grid in Digital Breast Tomosynthesis.  Radiological Society of North America 2013 Scientific Assembly and Annual Meeting, December 1 - December 6, 2013 ,Chicago IL. http://archive.rsna.org/2013/13014420.html