Figure 3. Image showing the vascular distances (left) and original image with the blood vessels masked out (right). Weighted addition of the information from these two images produces a combined image (middle), which allows more robust detection of lobar boundaries.