Abstract
BACKGROUND: To compare the doses calculated by the analytical anisotropic algorithm (AAA) and two dose reporting modes of Acuros XB (AXB(D(m)) and AXB(D(w))) with varied CT values on the Eclipse (Varian Medical Systems, Palo Alto, CA). MATERIALS AND METHODS: Virtual phantoms with a central layer of heterogeneous material (thickness = 2 or 5 cm) were created with Eclipse. Using single or opposed fields, the field sizes were 5 x 5 cm(2) or 10 x 10 cm(2). The photon energies were 6 or 10 MV, and the source-to-target distance was 100 cm. The relative doses at the center of the heterogeneous material layer were evaluated with varied CT values, from -1000 to 3000 HU. Values were normalized with the dose at 0 HU (100%) for comparative analysis. RESULTS: The results obtained from continuous data for a single field, 6 MV, 5 x 5 cm(2), and the heterogeneous material 5 cm, where the differences between algorithms were most pronounced, were as follows. In the low-density region (-1000 HU and -800 HU), the dose differences for AXB with reference to AAA were, respectively, -54.5% and +4.6% (AXB(D(m))) and -47.0% and +3.5% (AXB(D(w))), and in the high-density regions (1000 HU and 3000 HU) were -5.7% and -8.8% (AXB(D(m))) and +7.4% and +3.5% (AXB(D(w))), respectively. Consequently, dose differences at arbitrary CT values could be obtained. CONCLUSION: Dose differences between these algorithms were clarified for heterogeneous materials. The risk of dose reduction or escalation in clinical use was clearly visible between CT values from -1000 to 3000 HU.