Convolutional neural networks predict the linear energy transfer for proton-beam radiotherapy of patients with brain tumours

Sebastian Starke, Jan Eulitz, Alex Zwanenburg, Esther G.C. Troost, Mechthild Krause, Armin Lühr, Steffen Löck

Keywords: proton therapy, linear energy transfer, relative biological effectiveness, convolutional neural networks

Abstract: Proton therapy is a promising option for cancer treatment, even though its radiobiological properties are not yet fully considered in clinical practice. In this context, the relative biological effectiveness (RBE) of protons is the most important quantity, which is strongly related to their linear energy transfer (LET). LET distributions can be provided by commercial treatment-planning systems based on Monte Carlo simulations. However, such systems require a considerable amount of computational resources, are not yet available in every proton-therapy centre and may not be applicable to assess retrospective patient data. Here, we provide proof-of-concept for inferring LET distributions using convolutional neural networks (CNN) based on proton therapy radiation dose distributions and treatment-planning computed tomography (CT). We further evaluate established models for estimating treatment-related side effects after proton therapy of brain tumours and observe good agreement between CNN and MC based outputs.

Registration: I acknowledge that acceptance of this work at MIDL requires at least one of the authors to register and present the work during the conference.

Authorship: I confirm that I am the author of this work and that it has not been submitted to another publication before.

Paper Type: novel methodological ideas without extensive validation

Primary Subject Area: Application: Radiology

Secondary Subject Area: Application: Other

Confidentiality And Author Instructions: I read the call for papers and author instructions. I acknowledge that exceeding the page limit and/or altering the latex template can result in desk rejection.