Recent work proposed the use of magnetic field as a solution to reduce the undesirable effect of air cavities on dose after the air/tissue interface. In contrast to the published work that looks into the problem with slab geometries, in this work we use actual anatomy based on CT images and the magnetic flux from a Helmholtz coil-pair configuration to investigate the problem and to evaluate the efficacy of the proposed solution. The EGS4 phantom was created using CT scans of the head at the level of the ethmoid sinus. The sinus measures 1.95 x 2.18 x 2.00 cm3. The grid size used is 0.15 x 0.15 x 0.4 cm3. Three different radiation beams, 1 x 1, 2 x 2, and 4 x 4 cm2, all 6 MV irradiate the phantom in two different configurations: single beam and parallel opposed. The magnetic field has three different strengths: 0.0, 0.5, and 1.0 T. These represent the maximum strength achieved in the middle of the configuration, between the two coils. The depth of the second buildup region in the absence of the magnetic field was used as the normalization point for the purpose of analysis. Dose was then scored at 0.23 cm after the air/tissue interface. A second phantom, very similar to the CT-based phantom, was created, but with the sinus cavity filled with unit-density tissue; everything else remained the same. This phantom provides a base to investigate the effect of the air cavity on dose. The phantom was termed the phantom without air, or PWA for short. We use the terms "dose reduction ratio" (DRR), defined as one minus the ratio of the dose in PWA to the dose with the presence of the cavity multiplied by 100% and the "dose improvement ratio" (DIR), defined as the ratio of dose with B to that without B, to evaluate the reduction in dose due to the cavity and the improvement in dose with magnetic field, respectively. For single beam geometry, the reduced dose ranged from 41% (1 x 1 cm2 beam) to less than 2% (4 x 4 cm2 beam). For the same single beam geometry, DIR ranged from 1.13 to 1.00 (DIR = 1 indicates no change) with 0.5 T, whereas it ranged from 1.44 to 1.05 for 1.0 T magnets. When an opposing beam was used, the reduced dose was not as severe, such that DRR ranged from 24% to less than 2%. Whereas the dose improvement ranged from 1.08 to 1.00 for 0.5 T, and from 1.23 to 1.01 for 1.0 T.