Tomosynthesis, a technique that can produce pseudo 3-D reconstructions, could top traditional mammography in breast cancer screening and detection, a recent study suggests. Researchers at Emory University have found that digital tomosynthesis imaging offers more sensitivity and specificity than mammography in breast cancer screening. “Currently, mammography is the technique used most often… but since it gives only 2-D projection information of a 3-D anatomical structure, inaccuracies in screening often occur,” said James G. Nagy, a researcher and professor of mathematics and computer science at the university. Digital tomosynthesis imaging, which involves multiple projections of an object obtained along varying incident angles, can be used to reconstruct a pseudo 3-D representation of the object. These reconstructions are generated by computing weight fractions of the individual materials composing the object. Nagy’s group is currently developing an implementation with flexibility to exploit architecture as well as hybrid CPU/GPU (central processing unit/graphics processing unit) systems. “This novel reconstruction algorithm will allow us to perform tomosynthesis using more advanced image acquisition methods, which may result in further improvements in image quality, lower radiation dose and material decomposition,” Nagy said. The underlying idea is that several 2-D image projections taken at varying angles can provide more comprehensive and different information about the 3-D object. Conventional x-ray mammography is limited by superposition of breast tissue, which can sometimes mimic or obscure malignant pathology. The polyenergetic nature of the x-ray also is accounted for in the research. “All reconstruction algorithms used by clinical machines use a simple, but incorrect, assumption that the x-ray beam is monoenergetic,” Nagy said. “The polyenergetic model attempts to provide a more accurate physical model of the x-ray projection process, which results in more accurate reconstructions of images. The trade-off is that the mathematical problem is more complicated, and the computational costs are greater.” Future research will focus on algorithms that are computationally more economical. The research was published in the SIAM Journal on Scientific Computing. For more information, visit www.emory.edu.