Fundamentals of Breast Tomosynthesis Improving the Performance of Mammography

Theory of Tomosynthesis Conventional x-ray mammography is a two-dimensional imaging modality. In conventional mammography, pathologies of interest are sometimes difficult to visualize because of the clutter of signals from objects above and below. This is because the signal detected at a location on the film cassette or digital detector is dependent upon the total attenuation of all the tissues above the location. Tomosynthesis1,2,3,7,8 is a three-dimensional method of imaging that can reduce or eliminate the tissue overlap effect. While stabilizing the breast, images are acquired at a number of different x-ray source angles. Objects at different heights in the breast display differently in the different projections. In Figure 1, two objects (a spiculated lesion and ellipse) superimpose when the x-rays are at 0o, but the off-axis acquisitions shift the objects’ shadows relative to one another in the images. The final step in the tomosynthesis procedure is reconstructing the data to generate images that enhance objects from a given height by appropriate shifting of the projections relative to one another. An example is shown in Figure 2 where we reconstruct a cross sectional slice at one specific height. In this example, the images are summed, shifting one relative to another in a specific way that reinforces the spiculated lesion object and reduces the contrast of the ellipsoidal object by blurring it out. Note that additional acquisitions are not required to enhance the visibility of objects at any given height—one set of acquired data can be reprocessed to generate the entire 3D volume set.