Date Approved

8-8-2026

Graduate Degree Type

Project

Degree Name

Medical Dosimetry (M.S.)

Degree Program

School of Interdisciplinary Health

First Advisor

Vicente Ornelas

Second Advisor

Somayeh Gholami

Third Advisor

Kristen Vu

Academic Year

2025/2026

Abstract

Purpose

Spatially fractionated radiation therapy (SFRT) has demonstrated clinical benefit for bulky tumors; however, treatment planning remains challenging due to complex anatomy, proximity to critical organs at risk (OARs), and the lack of standardized VMAT-based workflows. Variations in contouring, lattice geometry, optimization, and multileaf collimator (MLC) design further limit reproducibility. This study developed and evaluated a practical VMAT-based lattice SFRT workflow for bulky head and neck tumors across multiple linear accelerator (linac) platforms with different MLC designs.

Methods

A workflow was developed using Eclipse TPS (v18.1, Acuros XB) with 6 MV-FFF beams. Ten retrospective bulky head and neck cases were planned on Varian Edge (HD120 MLC), TrueBeam (Millennium 120 MLC), and Halcyon (dual-layer MLC) systems. Single-fraction plans prescribed 15 Gy to lattice vertices while maintaining valley doses below 6 Gy. Spherical vertices (0.5 cm diameter) were placed 3–5 cm apart center-to-center. VMAT plans used 1–4 arcs with 10°–30° collimator rotations. Strategic normal tissue optimization emphasized OAR sparing and steep dose gradients. Plan quality was evaluated using dose-volume histograms, peak-to-valley dose ratio (PVDR), modulation factor, and OAR maximum doses.

Results

Clinically acceptable lattice plans were generated on all three platforms. Modulation factors remained below 6, and PVDR ranged from 2.8–5.0. Edge produced steeper dose gradients due to its finer MLC resolution, while TrueBeam and Halcyon achieved comparable dosimetry following minor vertex spacing adjustments. Workflow adaptations enabled consistent, deliverable SFRT plans across all platforms.

Conclusions

A standardized VMAT-based lattice SFRT workflow enables reproducible, high-quality treatment plans across multiple linac platforms while accounting for machine-specific MLC characteristics.

Available for download on Sunday, August 08, 2027

Share

COinS