Introduce the first SPArc optimization algorithm for synchrotron accelerator based proton therapy system.

BACKGROUND: Spot-scanning Proton Arc (SPArc) has drawn significant interest from the particle therapy community because of its superior target dose conformity, Organs-At-Risk (OARs) sparing, and Linear Energy Transfer (LET) modulation. However, all the existing SPArc optimization algorithms are designed for cyclotron-based proton treatment systems (PTS) instead of synchrotron-based PTS with the Single Energy Extraction (SEE) technique. The synchrotron-based PTSdelivers a bunch of proton particles through each cycle, and remaining proton particles will be discarded, resulting in significantly prolonged delivery time for the SPArc plan compared to PTS-cyclotron. PURPOSE: This study aims to develop the first SPArc optimization algorithm based on the dynamic programming (DP, SPArc-DP), to improve the delivery efficiency for synchrotron-based PTS. METHOD: Dynamic Programming was introduced to optimize the energy layer and the monitor unit (MU) distribution based on the features from the synchrotron-based PTS. It started from a plan generated via the original SPArc algorithm (SPArc-original) and iteratively merges the adjacent energy layers into the same energy layer while ensuring the plan quality. Thus, it effectively reduces unnecessary cycling. Four representative disease sites are selected for testing, including the liver, HN, lung and prostate cancers. Dosimetric metrics, including target performance and OARs sparing, were evaluated, and the delivery efficiency were simulated and compared between SPArc-original and SPArc-DP. RESULT: With similar plan quality, the SPArc-DP plans reduced acceleration cycles from 154.75 ± 22.70 counts to 54.25 ± 15.58 counts on average. This improvement effectively saved total treatment delivery time by 210.47 s ± 40.93 s (relatively 57.37% ± 6.96%) compared to SPArc-original plans.