Abstract of the talk// Adapting proton treatment plans is crucial to maintain tumor control and ensure optimal sparing of organs at risk when anatomy deviates from the planning CT and leads to significant dose degradation. Performing such adjustments while the patient is on the treatment couch — known as online adaptive proton therapy (OAPT) — maximizes the potential benefits of proton therapy, but an online adaptive workflow is challenging as all the complex tasks have to be done within a limited clinical timeframe. Therefore, integrating automated and accurate decision support plays a vital role in streamlining the OAPT workflow.

One of the first evaluation-based OAPT workflows in Europe shall be implemented at the University Proton Therapy Dresden. Thu and Jonathan are working on developing automated decision support systems performing different tasks.

In the first part of the talk, Thu will introduce the basic principles of proton therapy, the workflow of OAPT and its benefits. She will talk about information from different data domains arising during the treatment which can be assessed to determine the adaption need. The types of extracted features, and pros and cons of each domain will also be presented. Finally, she will discuss first results related to the features from occurring image changes during the treatment course of lung cancer patients.

In the second part of the talk, Jonathan will share insights to another project in the scope of OAPT: Treatment verification with a prompt gamma imaging slit camera. After introducing the basic principle and benefits of this technique, Jonathan will elaborate on the past and current steps towards its clinical implementation. In particular, different classification methods and their performance for the detection of treatment deviations will be demonstrated. As of now, the focus of the project is preparing for a clinical interventional study, which will employ an updated classification method to trigger online CT imaging in case of a detected treatment deviation. Jonathan will also highlight the challenges for the clinical translation of new technologies.