When you drive to an unfamiliar neighborhood, you don't need a map that shows every house. Streets and turns are enough. The same goes for calculating radiation dose in breast CT: you don't need every detail, just the big picture.
Previously, for each patient, a model of the breast was built from millions of tiny cubes (voxels). The computer had to process all of them, and the calculation took about 27 hours. That's like studying the floor plan of every apartment in every building before a trip.
It turned out that the size of the cube in the imaging plane affects the dose more than its thickness. So the thickness was increased more, while the width and height were increased less. As a result, the number of cubes dropped by a factor of 20–31, and the dose changed by only a few percent. Then they also reduced the number of simulated particles. The calculation now takes 72 minutes instead of 27 hours, with accuracy nearly unchanged.
This approach could be useful in hospitals: the dose can be estimated quickly, even before the main scan. The coarse map turned out to be just as good as the detailed one, but 22 times faster.
The original phantom of one breast contained up to 95 million voxels, and after optimization only about 3 million remained. Yet the difference in the calculated dose was less than half a percent of the target radiation level.