Room: Karl Dean Ballroom A1
Purpose: To assume the clinical benefits of radiation therapy, treatment optimization must be performed efficiently, in real time, and with minimal or no compromise to the original planning objectives. For complex techniques such as intensity-modulated radiation therapy (IMRT), this represents a challenge. Quantum annealing (QA) may be innately advantageous for complex optimization compared to traditional optimization techniques: its additional degree of freedom enhances its ability to explore the solution space and potentially find the global optimum within the annealing schedule while accounting for uncertainties. The aim of this study is to investigate the performance characteristics of the QA algorithm application to IMRT optimization.
Methods: A QA algorithm was implemented for beamlet-intensity IMRT optimization on a stereotactic body radiotherapy (SBRT) liver case. To compare with competing techniques, optimizations were also performed using a simulated annealing algorithm over the same parameters and number of iterations. The gold standard for assessing clinical feasibility consisted of dose volume histogram (DVH) metrics, which were obtained from plan optimizations generated using a commercial treatment planning software. After tuning the objective parameters to achieve clinically reasonable results, 50,000 iteration optimizations were repeated for each algorithm 200 times on the SBRT case to investigate each algorithmâ€™s sensitivity.
Results: Initial results on the SBRT case suggest QA exhibited greater flexibility in its â€˜walkâ€™ along the energy landscape and could converge to better solutions ~87% of the time. This difference translated into better coverage for the planned target volume by 22.8Â±0.98%.
Conclusion: These results suggest that the QA algorithm holds promise as a tool for achieving better treatment plans. An intriguing focus of future study is whether QAâ€™s additional parameter (the barrier width function) can be tuned to achieve maximally advantageous results across varying treatment cases.
Optimization, Inverse Planning, Intensity Modulation
TH- External beam- photons: IMRT dose optimization algorithms