Authors (including presenting author) :
Lee FKH (1), Kong CWY (1), Cheung FCH (1), Yip CWY (1), Chan SKY (1), Leung AKH (1), Choi VWY (1), Tsang TCK (1), Wong MKC (1)
Affiliation :
(1) Medical Physics Division, Clinical Oncology Department, Queen Elizabeth Hospital
Introduction :
Modulated radiotherapy is a highly complex treatment technique which requires fine coordination of mechanical motion (multi-leaf-collimators, gantry rotation, etc) and radiation output of linear accelerator (linac). According to international protocol on quality of radiotherapy, quality assurance (QA) must be conducted for each treatment plan to ensure high consistency of the delivered radiation dose compared to the original plan. In a busy department with 8 linacs, these QA require substantial linac machine hours for measurement. Oftentimes medical physics staff have to stay late to conduct QA after all patient treatments have finished. An efficient QA program is necessary to satisfy the increasing clinical demand.
Objectives :
The result of our QA program for the previous 3 years (Jan 2018 – Dec 2020) was evaluated, and the effectiveness was assessed.
Methodology :
A QA program for modulated radiotherapy was established in our department a few years ago, which included the use of independent dose calculation software as a type of ‘virtual QA’. This screened out treatment plans with a low possibility of failing a QA (e.g. plans without excessive modulation, and thus less likely to over-demand machine capability), and such cases can spare real measurement. This virtual QA was conducted only for selected types of plans, which were usually typical and less complicated cases. Moreover, for a linac recently installed or a planning system recently upgraded, we resorted to the ordinary measurement QA, until sufficient data was gathered to establish a screening threshold and that the machine performance has stabilized.
Result & Outcome :
QA was conducted on 3108 treatment plans over the period. 69 (2.22%) of the cases received a marginal pass result, while 7 (0.23%) failed. 1259 (40.51%) virtual QA was performed, out of which 238 (18.9%) failed and therefore require ordinary measurement QA. In total, 1021 real measurement QA was spared, saving over 250 linac machine hours. The trend revealed a higher rate of cases falling into marginal pass, after a major hardware replacement or upgrade was performed. The result was later improved after readjustment of hardware or beam data configuration. The quality of the modulated radiotherapy treatment plans was therefore maintained.
