Evaluation of fluence‐smoothing feature for three IMRT planning systems

Christopher J Anker,Bill J Salter,Dennis C Shrieve,Julie Chapek,Matt Tobler,Brian Wang,Ying J Hitchcock

doi:10.1120/jacmp.v11i2.3035

Christopher J Anker, Bill J Salter + Show 5 more

Open Access

https://doi.org/10.1120/jacmp.v11i2.3035

Copy DOI

Abstract

Commercially available intensity‐modulated radiation therapy (IMRT) inverse treatment planning systems (ITPS) typically include a smoothing function which allows the user to vary the complexity of delivered beam fluence patterns. This study evaluated the behavior of three ITPSs when varying smoothing parameters. We evaluated four cases treated with IMRT in our clinic: sinonasal carcinoma (SNC), glioblastoma multiforme (GBM), base of tongue carcinoma (BOT), and prostate carcinoma (PST). Varian Eclipse v6.5, BrainLAB BrainScan v5.31, and Nomos Corvus v6.2 ITPSs were studied for the SNC, GBM, and PST sites. Only Eclipse and Corvus were studied for BOT due to field size constraints of the BrainLAB MM3 collimator. For each ITPS, plans were first optimized using vendor‐recommended default “smoothing” values. Treatment plans were then reoptimized, exploring various smoothing values. Key metrics recorded included a delivery complexity (DC) metric and the Ian Paddick Conformality Index (IPCI). Results varied widely by vendor with regard to the impact of smoothing on complexity and conformality. Plans run on the Corvus ITPS showed the logically anticipated increase in DC as smoothing was decreased, along with associated improved organ‐at‐risk (OAR) sparing. Both Eclipse and BrainScan experienced an expected trend for increased DC as smoothing was decreased. However, this increase did not typically result in appreciably improved OAR sparing. For Eclipse and Corvus, and to a much lesser extent BrainScan, increases in smoothing decreased DC but eventually caused unacceptable losses in plan conformality. Depending on the ITPS, potential benefits from optimizing fluence smoothing levels can be significant, allowing for increases in either efficiency or conformality. Because of variability in smoothing function behavior by ITPS, it is important that users familiarize themselves with the effects of varying smoothing parameters for their respective ITPS. Based on the experience gained here, we provide recommended workflows for each ITPS to best exploit the fluence‐smoothing features of the system.PACS numbers: 87.56.bd, 87.56.N‐

Highlights

34 Anker et al.: Evaluation of intensity-modulated radiation therapy (IMRT) smoothing incorporate computer optimization approaches in the solution of this large scale problem
Assuming that “unnecessary noise” exists in the fluence map, this can allow for increased delivery efficiency without degradation of plan quality, as decreased fluence complexity requires fewer monitor units for dose delivery.[12] smoother intensity maps may result in dose distributions that are less sensitive to patient motion and treatment alignment uncertainties,(5,11) and plan quality analyses and treatment execution may be easier.[7]. For some treatment planning systems, these benefits may be seen with little to no compromise in plan quality and dose distribution.[5,6,8,11]
We explore the behavior of such smoothing functions in three commercially available inverse treatment planning systems (ITPS) (Eclipse-Varian Medical Systems, Palo Alto, CA; BrainScan, BrainLAB AG, Feldkirchen, Germany; and CORVUS, Best Nomos, Pittsburg, PA) for four challenging cases treated by IMRT in our clinic

Summary

Introduction

34 Anker et al.: Evaluation of IMRT smoothing incorporate computer optimization approaches in the solution of this large scale problem. Available inverse treatment planning systems (ITPS) typically include a “smoothing” or “efficiency” interface This allows the user to manipulate key parameters of the vendor-specific smoothing function that control the irregularity or complexity of delivered beam fluence patterns. We explore the behavior of such smoothing functions in three commercially available ITPSs (Eclipse-Varian Medical Systems, Palo Alto, CA; BrainScan, BrainLAB AG, Feldkirchen, Germany; and CORVUS, Best Nomos, Pittsburg, PA) for four challenging cases treated by IMRT in our clinic. This analysis does not seek to directly compare one ITPS to another, but rather to improve understanding of each ITPS’s smoothing algorithm by describing them in parallel

Methods

Results

Discussion

Conclusion