68-4

64/136

(a) original, (b) manual FIF, (c) FIF, and (d) FIF-4SF plans for this case. The FIF plan was rejected because the reduction in D95% was >1% (1.6%), and Step 2 was performed. Figure 9d shows the dose distribution in the FIF-4SF plan. In the dose distributions of the manual FIF and FIF plans (Figure 9b and c), a low-dose region appeared in the field edge of the sub-beam. As shown in the axial image in Figure 9c, the 100% dose region in the FIF plan was smaller than that in the original plan. In the dose distribution in FIF-4SF (Figure 9d), the dose reduction around the field edge of the sub-beam was small compared with the dose distri-bution in the FIF plan.384Comparing the DVHs of the manual FIF, FIF, and FIF-4SF plans in Figure 8b, the maximum dose region that corresponds to the tail part of the DVHs is nearly similar in the three FIF plans. By increasing the number of sub-beams from two to four in the FIF-4SF plan, D95% became close to the corresponding values in the original plan. The reduction of D95% from the original plan was 1.6% in the FIF plan, and 0.7% in the FIF-4SF plan.DiscussionIn this study, sub-beam creation and weight adjustment in FIF treatment planning were auto-mated. These are the time-consuming parts in the FIF treatment planning and require ≥6 min if performed manually, as shown in Figure 6. The automatic script can be started with one click and requires <2 min, and reduces the burden on the treatment planner. Comparing the semiautomatic FIF and manual FIF plans, there was no significant difference in the HI; there were significant differ-ences in V95% and D95%. However, the differences of V95% and D95% were small. Thus, we found that the semiautomatic planning script could create the FIF plans almost equivalent to the manual FIF plans.If a TPS has a scriptable interface, the proposed method can be easily introduced into the TPS. In this study, we adopted the change in D95% as the criterion for selecting the FIF technique. The crite-rion can be easily modified by adjusting the param-eters in the automatic FIF planning script, depending on the treatment site and protocol of each institu-tion. The threshold value for the hotspot regions and number of sub-beams can also be changed. For example, if Dth is lowered, both Dmax and D95% will decrease. If the reduction of D95% is acceptable, lowered Dth can be used. In this study, Dth was taken as 105% of the prescribed dose to make the maximum dose less than 107% of the prescribed dose considering the ICRU recommendation. Dth can be changed by considering a balance between Dmax and D95%. In this study, the beam energies of the sub-beams were the same as those of the main beams (10 MV). The use of different energies for the sub-beams can be easily implemented and may improve the dose distribution.Semiautomatic FIF planning was performed for the WBI cases, which were previously treated with the two-lateral-opposing-field technique. In the treatment plans with two-lateral-opposing tech-nique, hotspot regions appeared in the frontal and occipital lobes, which are laterally thin parts of the brain2, 4). Although there were significant differ-ences in V95% and D95% between the original and semiautomatic FIF treatment plans, the average decreases in V95% and D95% were 0.1% and 0.4%, respectively. The effect of the reductions in V95% and D95% is expected to cause minimal deterioration of the quality of the semiautomatic FIF plans.In this study, we adopted a 2-Step schema to select FIF technique. In the two-step FIF scheme, the number of sub-beams was increased from two to four to improve dose coverage of PTV, when the reduction in D95% was >1%. One case resulted in the FIF-4SF plan. The reduction of D95% compared with those in the original plan were 1.6% in the FIF plan. In the FIF-4SF plan, the reduction became 0.7% by increasing sub-beams. This indicates that the two-step FIF scheme can individualize the complexity of a treatment plan (beam number) depending on a patient.The monitor unit (MU) for sub-beams becomes small by increasing the number of sub-beams. In the FIF plans, the weight of the main beam and sub-beam on the average was 0.979 ± 0.010 and 0.021 ± 0.010, respectively. In the manual FIF plans, the weight of the main beam and sub-beam on the average was 0.975 ± 0.011 and 0.036 ± 0.046. In the FIF-4SF plan, the weights of the main beams, first sub-beams, and second sub-beams were 0.973, 0.013, and 0.014, respectively. Small MU may cause dose uncertainty. To avoid the dose uncertainty due to small MU, the algorithm to limit the minimum MU is easily implemented in the automatic FIF