Adaptive motion mapping in pancreatic SBRT patients using Fourier transforms

Area of Science:

• Medical Physics
• Radiation Oncology
• Image-Guided Therapy

Background:

• Four-dimensional computed tomography (4DCT) has limitations in accurately measuring respiratory-induced pancreatic tumor motion.
• Daily variations in pancreatic tumor movement during stereotactic body radiation therapy (SBRT) require advanced assessment.
• Pre-treatment imaging may not fully capture the dynamic nature of tumor motion.

Purpose of the Study:

• To assess the daily motion of pancreatic tumors during SBRT.
• To develop adaptive strategies for predicting and accounting for pancreatic tumor motion.
• To evaluate a novel metric for classifying respiratory motion characteristics.

Main Methods:

• Utilized a model to reconstruct instantaneous 3D tumor position from 2D CBCT projection images.
• Developed a "Spectral Coherence" (SC) metric based on the Fourier frequency spectrum of motion.
• Analyzed SC's ability to predict motion errors and classify patient motion patterns.

Main Results:

• Daily tumor motion amplitude exceeded 4DCT predictions by up to 3.5mm in different directions.
• SC correlated with daily motion variations and affected tumor dose coverage.
• An adaptive protocol using SC significantly improved target dose metrics (D95, D99, minimum dose).

Conclusions:

• A Fourier-based approach effectively differentiates respiratory motion consistency.
• The SC metric correlates with daily motion deviations from 4DCT.
• An SC-based adaptive protocol demonstrated feasibility for improving pancreatic SBRT dosimetry by tailoring margins to patient-specific motion.