Four-dimensional dosimetry validation and study in lung radiotherapy using deformable image registration and Monte...

Area of Science:

• Medical Physics
• Radiation Oncology
• Computational Biology

Background:

• Thoracic cancer treatment is complicated by respiratory motion and lung density variations, impacting radiation dose accuracy.
• Four-dimensional (4D) dose calculations using Monte Carlo methods and deformable image registration are proposed solutions.

Purpose of the Study:

• To validate 4D Monte Carlo dosimetry against measurements.
• To compare 4D dosimetry for varying tumor sizes and motion ranges.
• To assess the impact of respiratory phase inclusion on dose-volume histograms (DVH).

Main Methods:

• Utilized BEAMnrc for dose calculations and an optical flow algorithm for deformable image registration and dose mapping.
• Validated calculations using a moving phantom, comparing measured and calculated doses within gross tumor volumes (GTV).
• Analyzed 4D CT datasets from lung cancer patients.

Main Results:

• Calculated and measured doses in a moving phantom agreed within 3% at the GTV center.
• Small tumors with significant motion (1.5 cm) showed reduced tumor coverage with a 1 cm beam margin in 4D dosimetry.
• Large tumors or those with minimal motion (<1 cm) showed no significant difference between 4D and static plans.
• DVH analysis indicated that using only end-expiration and end-inspiration phases approximates all-phase calculations for similar lung cancer cases, reducing computational load.

Conclusions:

• 4D Monte Carlo dosimetry is a validated and accurate method for thoracic cancer radiation therapy.
• Tumor size and motion significantly influence the necessity and impact of 4D dosimetry.
• Simplified 4D dosimetry using extreme respiratory phases is a viable approach for lung cancer, optimizing calculation efficiency.