The idea is to develop and evaluate a quantum-inspired formulation for optimization problems encountered in intensity-modulated radiation therapy (IMRT) for cancer treatment. Quantum computing has the unique ability to consider the entire optimization parameter space simultaneously, which would greatly accelerate the optimization process, enable real-time planning and ensure that the best solution is found, optimizing radiation therapy for patients. However, quantum hardware is years away from being feasible.
A quantum-inspired formulation, as proposed in this project, will develop a quantum-ready pathway to quantum computing utilization down the road as quantum hardware becomes clinically available. Furthermore, by using conventional computers, the development of a quantum-inspired formulation can have a significant impact on the quality and efficiency of clinical external beam radiation therapy planning today for both photon and proton treatments.
Secondary to the limitations of conventional computers, treatment planning for IMRT and intensity-modulated proton therapy (IMPT) cannot be done in real-time to account for changes in patient anatomy, limiting the opportunity to further improve outcomes for cancer patients. Ultrafast computing can make real-time planning clinically feasible, allowing most patients to benefit from more accurate delivery with less toxicity. Another challenge in IMRT/IMPT planning is that the traditional optimization algorithm often is trapped in a local minimum rather than the global minimum. In many cases, a more optimal treatment plan exists that would further minimize radiation to at-risk organs, which would improve patient outcomes and the therapeutic ratio of cancer care.
Our first goal is to develop a streamlined process for practical usage and evaluation of our quantum-inspired algorithms using conventional computers to perform mathematical optimization. The second goal is to apply quantum-inspired algorithms for proton beam angle selection and energy selection, which will well prepare us for clinical implementation of proton radiotherapy in the future.
We will test 20 breast cancer cases and compare the optimization solutions with the clinical solution in both speed and plan quality. We will test 35 breast and brain tumor cases and compare the proton beam angles and energy selections with the empirically selected beam angles and energies. Our prior work indicates that a quantum-inspired formulation using ordinary computers offers the potential for clinically oriented innovation that can improve patient outcomes over conventional radiotherapy treatment planning methods. In the long term, our “quantum-ready” algorithm can be implemented on future quantum hardware when feasible for clinical use.
100% For The Cure
