Experimental verification of the Acuros XB and AAA dose calculation adjacent to heterogeneous media for IMRT and RapidArc of nasopharygeal carcinoma

Research output: Journal Publications and ReviewsRGC 21 - Publication in refereed journalpeer-review

55 Scopus Citations
View graph of relations



Original languageEnglish
Article number31714
Journal / PublicationMedical Physics
Issue number3
Publication statusPublished - Mar 2013


Purpose: To compare the doses calculated by the Acuros XB (AXB) algorithm and analytical anisotropic algorithm (AAA) with experimentally measured data adjacent to and within heterogeneous medium using intensity modulated radiation therapy (IMRT) and RapidArc® (RA) volumetric arc therapy plans for nasopharygeal carcinoma (NPC). Methods: Two-dimensional dose distribution immediately adjacent to both air and bone inserts of a rectangular tissue equivalent phantom irradiated using IMRT and RA plans for NPC cases were measured with GafChromic® EBT3 films. Doses near and within the nasopharygeal (NP) region of an anthropomorphic phantom containing heterogeneous medium were also measured with thermoluminescent dosimeters (TLD) and EBT3 films. The measured data were then compared with the data calculated by AAA and AXB. For AXB, dose calculations were performed using both dose-to-medium (AXB-Dm) and dose-to-water (AXB-Dw) options. Furthermore, target dose differences between AAA and AXB were analyzed for the corresponding real patients. The comparison of real patient plans was performed by stratifying the targets into components of different densities, including tissue, bone, and air. Results: For the verification of planar dose distribution adjacent to air and bone using the rectangular phantom, the percentages of pixels that passed the gamma analysis with the ± 3%/3mm criteria were 98.7%, 99.5%, and 97.7% on the axial plane for AAA, AXB-Dm, and AXB-Dw, respectively, averaged over all IMRT and RA plans, while they were 97.6%, 98.2%, and 97.7%, respectively, on the coronal plane. For the verification of planar dose distribution within the NP region of the anthropomorphic phantom, the percentages of pixels that passed the gamma analysis with the ± 3%/3mm criteria were 95.1%, 91.3%, and 99.0% for AAA, AXB-Dm, and AXB-D w, respectively, averaged over all IMRT and RA plans. Within the NP region where air and bone were present, the film measurements represented the dose close to unit density water in a heterogeneous medium, produced the best agreement with the AXB-Dw. For the verification of point doses within the target using TLD in the anthropomorphic phantom, the absolute percentage deviations between the calculated and measured data when averaged over all IMRT and RA plans were 1.8%, 1.7%, and 1.8% for AAA, AXB-Dm and AXB-D w, respectively. From all the verification results, no significant difference was found between the IMRT and RA plans. The target dose analysis of the real patient plans showed that the discrepancies in mean doses to the PTV component in tissue among the three dose calculation options were within 2%, but up to about 4% in the bone content, with AXB-Dm giving the lowest values and AXB-Dw giving the highest values. Conclusions: In general, the verification measurements demonstrated that both algorithms produced acceptable accuracy when compared to the measured data. GafChromic® film results indicated that AXB produced slightly better accuracy compared to AAA for dose calculation adjacent to and within the heterogeneous media. Users should be aware of the differences in calculated target doses between options AXB-Dm and AXB-Dw, especially in bone, for IMRT and RA in NPC cases. © 2013 American Association of Physicists in Medicine.

Research Area(s)

  • Acuros XB algorithm, anthropomorphic phantom, GafChromic EBT3 film, heterogeneous media, intensity modulated radiotherapy, RapidArc

Citation Format(s)