Digital breast tomosynthesis (DBT) is an extension of conventional full field digital mammography (FFDM) that produces quasi three-dimensional reconstructed images of the breast from a set of low-dose x-ray projections acquired over a limited angular range. As a consequence, DBT can overcome the main limitation of FFDM: tissue superposition.
The use of DBT in the clinical setting for breast cancer imaging is continuously growing throughout medical services worldwide, due to many studies showing that DBT provides a higher sensitivity and specificity for breast cancer detection than FFDM. But on top of that, there are several large-population based screening trials that have proved that DBT should be regarded as a better mammogram, and considered as the new breast cancer screening technique for the near future.
However, DBT potential is still undergoing a continuous evolution, with special focus in new reconstruction techniques, synthetic mammograms, and acquisition protocols that can enhance its clinical performance and save radiation dose to the patients.
Example of a malignant lesion (an in situ ductal carcinoma), not visible on standard digital mammography, but clearly defined on breast tomosynthesis.
In the AXTI group, we work on both the clinical and technical optimizations of DBT, in order to assess and increase its diagnostic performance within our hospital and beyond.
On one side, we work closely with radiologists and we perform reader studies to determine what is the most optimum way of using DBT in the clinic, by analyzing different reading workflows as well as state-of-the-art reconstruction algorithms and the role of synthetic mammograms on the clinical practice. Regarding our technical work, we also develop new image processing and reconstruction techniques aiming to improve the image quality of DBT, such as X-ray scatter correction and automatic segmentation algorithms. Other areas of research for DBT are breast compression force reduction, breast modeling, and breast dosimetry on DBT.