Automation of radiation treatment planning : Evaluation of head and neck cancer patient plans created by the Pinnacle3scripting and Auto-Planning functions

Speer S, Klein A, Kober L, Weiss A, Yohannes I, Bert C (2017)

Publication Type: Journal article

Publication year: 2017

Journal

Strahlentherapie und Onkologie Urban & Vogel

Book Volume: 193

Pages Range: 656-665

Journal Issue: 8

DOI: 10.1007/s00066-017-1150-9

Abstract

BACKGROUND: Intensity-modulated radiotherapy (IMRT) techniques are now standard practice. IMRT or volumetric-modulated arc therapy (VMAT) allow treatment of the tumor while simultaneously sparing organs at risk. Nevertheless, treatment plan quality still depends on the physicist's individual skills, experiences, and personal preferences. It would therefore be advantageous to automate the planning process. This possibility is offered by the Pinnacle3treatment planning system (Philips Healthcare, Hamburg, Germany) via its scripting language or Auto-Planning (AP) module. MATERIALS AND METHODS: AP module results were compared to in-house scripts and manually optimized treatment plans for standard head and neck cancer plans. Multiple treatment parameters were scored to judge plan quality (100 points = optimum plan). Patients were initially planned manually by different physicists and re-planned using scripts or AP. RESULTS AND DISCUSSION: Script-based head and neck plans achieved a mean of 67.0 points and were, on average, superior to manually created (59.1 points) and AP plans (62.3 points). Moreover, they are characterized by reproducibility and lower standard deviation of treatment parameters. Even less experienced staff are able to create at least a good starting point for further optimization in a short time. However, for particular plans, experienced planners perform even better than scripts or AP. Experienced-user input is needed when setting up scripts or AP templates for the first time. Moreover, some minor drawbacks exist, such as the increase of monitor units (+35.5% for scripted plans). CONCLUSION: On average, automatically created plans are superior to manually created treatment plans. For particular plans, experienced physicists were able to perform better than scripts or AP; thus, the benefit is greatest when time is short or staff inexperienced.

Authors with CRIS profile

Christoph Bert Professur für Med. Strahlenphysik

Involved external institutions

Rinecker Proton Therapy Center DE Germany (DE) EKS Engineering GmbH DE Germany (DE) Strahlentherapie Tauber-Franken DE Germany (DE)

How to cite

APA:

Speer, S., Klein, A., Kober, L., Weiss, A., Yohannes, I., & Bert, C. (2017). Automation of radiation treatment planning : Evaluation of head and neck cancer patient plans created by the Pinnacle3scripting and Auto-Planning functions. Strahlentherapie und Onkologie, 193(8), 656-665. https://doi.org/10.1007/s00066-017-1150-9

MLA:

Speer, Stefan, et al. "Automation of radiation treatment planning : Evaluation of head and neck cancer patient plans created by the Pinnacle3scripting and Auto-Planning functions." Strahlentherapie und Onkologie 193.8 (2017): 656-665.

BibTeX: Download