Skip to content
    Guidelines & Consensus

    Dynamic modulated arc therapy framework co-optimizes plan quality and delivery time

    Li et alJ Appl Clin Med Phys

    Conventional Volumetric Modulated Arc Therapy (VMAT) optimization generally treats delivery time as an emergent byproduct of control-point-centric models, often neglecting finite axis acceleration and machine dynamic limits. Dynamic Modulated Arc Therapy (DMAT) was introduced as a time-aware, modulation-steered planning framework that explicitly co-optimizes dosimetric plan quality, treatment delivery time, and modulation complexity prior to plan delivery.
    The DMAT architecture integrates direct machine emulation accounting for dynamic limits and axis synchronization, dynamic modulation control, and metric-based cost functions. Planning involves a user-selected modulation level ranging from -3 to +3 that simultaneously controls leaf-travel allowance, control-point (CP) density, total monitor units (MUs), and aperture complexity penalties. The optimization loop employs progressive resolution alternating dosimetric updates with deliverability sequencing, followed by dynamic non-uniform CP redistribution based on combined geometric and complexity parameters.
    DMAT performance was tested using a model of an accelerated C-arm delivery system operating at maximum specifications of 2.5 RPM gantry speed, 6.25 cm/s MLC leaf speed, and 3000 MU/min dose rate across representative head-and-neck, lung stereotactic body radiation therapy (SBRT), and prostate SBRT cases. Evaluated across disease sites, higher modulation levels led to increased aperture complexity and longer delivery times, but the resulting dosimetric benefits varied significantly by anatomical site. Head-and-neck cases demonstrated marked plan-quality enhancements with higher angular resolution and modulation, whereas prostate and lung SBRT showed diminishing returns beyond baseline settings. DMAT exposes explicit machine-aware timing during the planning process, offering a navigable trade-off space for time-sensitive workflows such as adaptive radiotherapy and motion-managed treatments.

    Key Takeaways

    • Evaluated on a simulated accelerated C-arm linac (2.5 RPM max gantry speed, 6.25 cm/s max MLC speed, 3000 MU/min max dose rate), DMAT demonstrated site-dependent quality-time trade-offs, providing substantial plan-quality gains for head-and-neck treatments while showing minimal incremental quality improvements beyond baseline for prostate and lung SBRT.
    • Prioritizing delivery efficiency via negative modulation levels (-3 to +3 scale) predictably shortened treatment delivery times by reducing total monitor units and aperture complexity, accompanied by quantifiable reductions in plan quality.
    Read the original sourceJ Appl Clin Med Phys