A phantom study comparing technical image quality of five breast tomosynthesis systems

Abstract

BackgroundDigital breast tomosynthesis (DBT) is a three-dimensional breast imaging method. DBT vendors employ various approaches in both image acquisition and data processing, which may affect image quality and radiation exposure to patients. ObjectiveThis study aimed to evaluate the performance of five DBT systems: Fujifilm Amulet Innovality (using both a standard mode and high-resolution mode), GE Senographe Essential, Hologic Selenia Dimensions, Planmed Clarity 3D, and Siemens Mammomat Inspiration. Materials and methodsThe performance of each device and imaging technique was evaluated and compared by phantom measurements performed with four quality assurance phantoms. Technical image quality assessments consisted of measuring artefact extent, in-plane resolution, relative noise power spectrum, and geometric accuracy. ResultsArtefact spreading varied remarkably between the devices, and the full width at half maximum values of artefact spread functions varied from 3.5 mm to 10.7 mm. Noticeable in-plane resolution anisotropy, determined using modulation transfer function (MTF) analysis, was typically observed between tube travel direction and chest wall-nipple direction. The MTF50 varied from 1.1 mm−1 to 1.6 mm−1 and from 1.5 mm−1 to 4.1 mm−1 in the tube travel and chest wall-nipple directions, respectively. Moreover, distinctly different noise power spectra were observed between the systems. The geometric accuracy in every system was within 0.5%. ConclusionTechnical image quality assessments with image quality phantoms revealed remarkable differences in artefact spread, in-plane resolution, and noise properties between the DBT systems and imaging methods.