How to read a digital breast tomosynthesis

The interpretation of a tomosynthesis requires a three-dimensional approach that takes into account both the data relative to the slice and the evolution of these data in the stack of slices. The main output of tomosynthesis is shape data and requires density contrast between the lesion and the surrounding tissue. The best results of tomosynthesis are obtained in breasts with a fat component. A shape anomaly found on a section will be characterized by its evolution on adjacent sections in order to differentiate the fibroconnective trabeculae from spiculated and converging lesions. The interpretation must then be dynamic by navigating back and forth in the space occupied by the anomaly. Two orthogonal tomosynthesis might be needed to characterize some complex anomalies. (C) 2018 Published by Elsevier Masson SAS.