Probabilistic evaluations to improve treatment plan robustness against uncertainties for intensity-modulated proton therapy

Intensity-Modulated Proton Therapy (IMPT) is an advanced cancer treatment that uses protons to deliver radiation more precisely to tumors while sparing surrounding healthy tissues. This precision makes IMPT an attractive treatment modality, but it also introduces challenges, as it is more sensitive to uncertainties such as patient positioning and variations in proton range. These uncertainties can affect the accuracy of tumor targeting and, consequently, compromise the effectiveness of the treatment.

To manage these uncertainties, it is essential to evaluate how treatment plans perform under various patient scenarios, assessing the robustness of the treatment plan. Probabilistic robustness evaluations (PREs) can accurately quantify this robustness by sampling and calculating the dose across a vast number of patient scenarios.

In this thesis, we introduced PREs using Polynomial Chaos Expansion to analyze dose trade-offs between the tumor and healthy tissue in 100,000 scenarios per patient. We evaluated current clinical practice of IMPT not only at our institution but also at a national and multi-institutional level, and we compared its performance against conventional radiotherapy.