DOS 542 - Week 4 Discussion
Initial Post: QA of Brachytherapy and Treatment Planning Systems
Quality assurance (QA) procedures for treatment planning systems (TPS) are difficult to design, because so much of what happens in a TPS is invisible to the end-user. Ensuring that a TPS is functioning correctly is critical to the proper operation of any radiation therapy program, since almost all plan dose calculations are performed by the TPS. Since the actual process of dose calculation is difficult to test, the best way to verify the proper operation is to perform a full range of tasks and verify that the same or similar results are achieved each time. In some cases, results can be checked against a benchmark. Calculated dose can be compared against delivered dose, or measurements of known geometries or CT densities can be compared against known values.
At Loyola, commissioning of a TPS is an extensive process that involves gathering data on scan parameters, beam models, output factors, and so on. The commissioning process is beyond the scope of this QA discussion. In order to verify that the TPS is staying within the specifications defined during commissioning, several tests can be performed at various intervals.
On a daily basis, individual plans can be checked against another piece of software that independently calculates monitor units (MU). Loyola uses RadCalc for second checks of every beam/field.1 The values calculated by both systems must agree within 3% in order to pass. Another form of daily QA for intensity modulated radiation therapy (IMRT) plans is a physical check with a scanning phantom device that records a map of intensity delivered to the surface of the device and compares it to dose calculated by the TPS.
On a monthly basis, various dose calculation and image processing tests are performed.2 The monthly procedure at Loyola involves recalculating dose on four different past patients, representing a range of typical treatment planning scenarios. The four patients are a rectal patient treated with a 3D conformal plan and wedges, an esophageal case treated with RapidArc, a head & neck case treated with IMRT, and a breast boost case treated with electrons. Every month, these four patients are loaded and their plans are recalulated. In order to evaluate the dose calculation algorithm, the maximum, minimum, and mean dose to several tissue types are measured and compared to baseline values. The structures evaluated are:
Rectal plan (3D with wedges):
Esophagus plan (RapidArc):
Head & Neck plan (static IMRT):
Breast Boost plan (electrons):
In addition to testing the doses calculated for each of these structures in each of these plans, CT density and spatial integrity are also checked by opening the same phantom scan each time and remeasuring the densities of several phantom inserts of known density. The values for water, bone, exhaled lung, and inhaled lung are compared against baseline values. The physical dimensions of the phantom scan in the X, Y, and Z dimensions are also measured and compared against known values.
On an annual basis, output calculated by the treatment planning system is compared against output measured in a water phantom at the linear accelerator.3 Measurements are taken at a variety of depths, field sizes, beam energies, and beam types. The goal is to ensure that depth dose curves for every scenario are still accurately producing appropriate plans. A more thorough examination of imaging is also peformed. CT densities for four phantom plugs are measured, and the plugs are also contoured. Once contours are created, their physical dimensions are measured and their volumes are calculated, and these values are compared against known values.
In most cases, the TPS operates as expected as long as the software has not been updated since the last test. For this reason, the version number of every component is also recorded at each QA interval.
- Roeske J. Treatment planning QA program. [Departmental Memorandum]. 2015.
- Eclipse monthly QA. [Excel Document]. 2015
- Roeske J. Annual Eclipse planning system QA procedures. [Departmental Memorandum]. 2015.