How should we model and evaluate breathing interplay effects in IMPT?

Area of Science:

• Medical Physics
• Radiation Oncology
• Computational Biology

Background:

• Breathing motion during Intensity Modulated Proton Therapy (IMPT) introduces interplay effects, complicating accurate dose delivery.
• Assessing these effects requires statistical methods that capture breathing variability for robust clinical evaluation.

Purpose of the Study:

• To develop and present a statistical method for modeling intra-fraction respiratory motion in IMPT.
• To assess clinical aspects of interplay evaluation, including irregular breathing, sensitivity to breathing changes, and required statistical power.

Main Methods:

• Comparison of two data-driven methods for generating artificial patient-specific breathing signals: sinusoidal and deep learning models.
• Investigation of the relationship between interplay doses and breathing parameters, analyzing sensitivity to small breathing variations.
• Application of the statistical method to analyze the interplay robustness of 4DCT and Internal Target Volume (ITV) plans in lung cancer patients.

Main Results:

• Deep learning models generate more realistic breathing signals than sinusoidal models.
• Small changes in breathing period significantly impact dose distribution, revealing a highly fluctuating interplay dose-parameter relationship.
• Limited sampling for interplay statistics introduces greater error than using sinusoidal models or ignoring breathing hysteresis.

Conclusions:

• The developed statistical method provides a trustworthy quantification of interplay effects in IMPT.
• 4DCT plans demonstrate better interplay robustness compared to ITV plans, which systematically fail robustness requirements even with 33 fractions.