Abstract: Tomosynthesis reconstructs a 3D object from a scan consisting of a limited number of projections. Hence, tomosynthesis requires much less radiation dosage as compared to computed tomography (CT). A major problem with tomosynthesis is image artifacts associated with incompleteness of data. In this paper, we propose a tomosynthesis approach to achieve higher image quality in a region of interest (ROI) than competing techniques. First, a low-resolution global CT scan is acquired. Then, a high-resolution local scan is performed with respect to the ROI. Finally, images of the ROI are reconstructed from these two datasets. Our numerical simulation results show that images of the ROI obtained by our approach are significantly better than the counterparts without using the global scan information.