Reducing inter- and intra-planner variability in radiotherapy plan output with a commercial knowledge-based planning solution. (September 2018)

• Record Type: Journal Article
• Authors: Scaggion, Alessandro; Fusella, Marco; Roggio, Antonella; Bacco, Simonetta; Pivato, Nicola; Rossato, Marco Andrea; Peña, Lucia Mariel Arana; Paiusco, Marta
• Is Part Of: Physica medica. Volume 53(2018)
• Page Start: 86
• Page End: 93
• Publication Date: 2018-09
• Subjects: VMAT -- Knowledge-based planning -- Planning quality -- Consistency -- Planning experience; Medical physics -- Periodicals; Biophysics -- Periodicals; Biophysics -- Periodicals; Imagerie médicale -- Périodiques; Radiothérapie -- Périodiques; Rayons X -- Sécurité -- Mesures -- Périodiques; Physique -- Périodiques; Médecine -- Périodiques; 610.153
• DOI: 10.1016/j.ejmp.2018.08.016
• Languages: English
• ISSNs: 1120-1797

Abstract

Highlights: Well-controlled experiment design: manual and RapidPlan-assisted planning are compared on the same patient cohort across 7. Planners with different level of planning experience. Plan Quality assessment on the basis of many DVH-related metrics and overall plan quality. Plan quality consistency measurement, both inter-planner and intra-planner. Abstract: Purpose: This study measured to which extent RapidPlan can drive a reduction of the human-caused variability in prostate cancer treatment planning. Methods: Seventy clinical prostate plans were used to train a RapidPlan model. Seven planners, with different levels of planning experience, were asked to plan a VMAT treatment for fifteen prostate cancer patients with and without RapidPlan assistance. The plans were compared on the basis of target coverage, conformance and OAR sparing. Inter-planner and intra-planner variability were assessed on the basis of the Plan Quality Metric formalism. Differences in mean values and InterQuartile Ranges between patients and operators were assessed. Results: RapidPlan-assisted plans matched manual planning in terms of target coverage, homogeneity, conformance and bladder sparing but outperformed it for rectum and femoral heads sparing. 8 out of 15 patients showed a statistically significant increase in overall quality. Inter-planner variability is reduced in RapidPlan-assisted planning for rectum and femoral heads while bladder variability was constant. The inter-planner variability of the overall plan quality, IQR of PQM%, was approximately halved for all patients. RapidPlan assistance induced a larger increase in plan quality for less experienced planners. At the same time, a reduction in intra-planner variability is measured with a significant overall reduction. Conclusions: The assistance of RapidPlan during the optimization of treatments for prostate cancer induces a significant increase of plan quality and a contextual reduction of plan variability. RapidPlan is proven to be a valuable tool to leverage the planning skills of less experienced planners ensuring a better homogeneity of treatment plan quality.