Session Item

Anatomical changes may greatly impact the delivered radiotherapy dose in a negative way like tumor underdosage or organs at risk overdosage. Treatment plans can be adapted to the anatomy of the day right before starting the treatment, either by shifting the couch by matching anatomies from CBCT with planning CT, or by deformable image registration or re-contouring and re-optimizing the treatment plan. However, this cannot account for in-treatment anatomical changes, such as respiratory-induced motion. This issue has been tackled by various clinical trials on real-time adaptations such as using gating (switching the treatment beam on and off depending on target location), MLC-tracking (moving the treatment beam to where the target is) or couch-tracking (moving the patient in the opposite direction of internal motion to keep the target in place). However, these solutions are geometric and instantaneous in nature, thus unable to solve suboptimally delivered dose earlier in the treatment fraction.

A more clinically relevant metric than tumor position for plan adaptation is the delivered dose.

In this presentation, I will give a brief overview of the literature on calculating the dose in real-time taking time-resolved anatomical changes into account, and ideas on how to act in real-time based on the dose. This will cover calculating the dose with simplistic approaches, EPID-based solutions and Monte Carlo solutions, taking real-time anatomical changes into account. The ideas for adaptation will cover dose-guided gating, MLC and couch tracking, couch corrections (by estimated final dose for numerous couch-shift scenarios) and plan re-optimizing on-the-fly, all with and without dose-repair functionalities.