Experimental validation of the Eclipse AAA algorithm

Abstract

The present study evaluates the performance of a newly released photon-beam dose calculation algorithm that is incorporated into an established treatment planning system (TPS). We compared the analytical anisotropic algorithm (AAA) factory-commissioned with "golden beam data" for Varian linear accelerators with measurements performed at two institutions using 6-MV and 15-MV beams. The TG-53 evaluation regions and criteria were used to evaluate profiles measured in a water phantom for a wide variety of clinically relevant beam geometries. The total scatter factor (TSF) for each of these geometries was also measured and compared against the results from the AAA. At one institute, TLD measurements were performed at several points in the neck and thoracic regions of a Rando phantom; at the other institution, ion chamber measurements were performed in a CIRS inhomogeneous phantom. The phantoms were both imaged using computed tomography (CT), and the dose was calculated using the AAA at corresponding detector locations. Evaluation of measured relative dose profiles revealed that 97%, 99%, 97%, and 100% of points at one institute and 96%, 88%, 89%, and 100% of points at the other institution passed TG-53 evaluation criteria in the outer beam, penumbra, inner beam, and buildup regions respectively. Poorer results in the inner beam regions at one institute are attributed to the mismatch of the measured profiles at shallow depths with the "golden beam data." For validation of monitor unit (MU) calculations, the mean difference between measured and calculated TSFs was less than 0.5%; test cases involving physical wedges had, in general, differences of more than 1%. The mean difference between point measurements performed in inhomogeneous phantoms and Eclipse was 2.1% (5.3% maximum) and all differences were within TG-53 guidelines of 7%. By intent, the methods and evaluation techniques were similar to those in a previous investigation involving another convolution-superposition photon-beam dose calculation algorithm in another TPS, so that the current work permitted an independent comparison between the two algorithms for which results have been provided.

Figure 1

The outline of the mantle field used in test case 5, including the block shape. This field was used to measure the central‐axis inline and crossline profiles at each depth (see text). The off‐axis profiles were measured along the lines with the arrows each at 12 cm distance from the central axis.

Figure 2

The outline of the multileaf collimator field used in test case 10, including the jaw settings at $\mathrm{X}1=3$ cm, $\mathrm{X}2=12$ cm, $\mathrm{Y}1=3$ cm, and $\mathrm{Y}2=19$ cm. This field was used to measure the central‐axis inline and crossline profiles at each depth (see text). The off‐axis profiles were measured along the lines with the arrows: inline at 5 cm and crossline at 4 cm distances.

Figure 3

Locations of the test points within the CIRS phantom (CIRS, Norfolk, VA). Source‐to‐surface distance is 100 cm, field size is $26\times 14$ cm, 500 monitor units. The yellow background is the tissue‐equivalent material.

Figure 4

Isodose distribution and location of the test points within the anthropomorphic phantom for the lung test case. $\text{SSD}=100$ cm, $12\times 28$‐cm field, 15 MV, 275 monitor units.

Figure 5

Isodose distributions and location of the test points within the anthropomorphic phantom for the neck test case. (a) Slice 8, 2.5 cm superior to the beam‐entry plane, and (b) slice 9, containing the beam‐entry point. Source‐to‐surface distance is 100 cm, field size is $10\times 16$ cm, 6‐MV beam, 275 monitor units.

Figure 6

The crossplane profile from golden beam data is shown, together with the measured data from the Tom Baker Cancer Centre (TBCC) and Cross Cancer Institute (CCI) for a 6 MV $25\times 25$‐cm open field at 4 cm depth. The line indicates the location of the off‐axis profile at 80% field edge.

Figure 7

Graphs showing the penumbra portions of measured (blue) and calculated (green) profiles in (a) 6‐MV and (b) 15‐MV beams at 10 cm depth for a $10\times 10$‐cm field. The other two curves in these graphs indicate the lower (red) and upper (magenta) limits according to penumbra criteria given in Table 2. Notice that the Eclipse data show a steeper penumbra than do the measured profiles at both energies. This general trend was observed at the other depths investigated in the present study.

Figure 8

The description of (a) calculated and (b) measured penumbrae of a $4\times 4$‐cm field is shown for three different cases: using a diamond detector; using an CC13 chamber; and data present in the AAA beam configuration denoted as GBD. The calculated data in the upper‐panel curves are obtained from Eclipse after the AAA is configured with the data measured by various detectors. The insensitivity of the modeled penumbra to the measured penumbra is clearly visible in (a). All curves show percentage dose relative to a $10\times 10$‐cm field at maximum dose $(D_{\text{max}})$.