Algorithm for correcting optimization convergence errors in Eclipse

• Published in: Journal of applied clinical medical physics Vol. 10; no. 4; pp. 281 - 289
• Main Authors: Zacarias, Albert S., Mills, Michael D.
• Format: Journal Article
• Language: English
• Published: United States John Wiley & Sons, Inc 14.10.2009; John Wiley and Sons Inc
• Subjects: AAA; Accuracy; Algorithms; Child; Humans; iterative correction; Lung Neoplasms - pathology; Lung Neoplasms - radiotherapy; Neuroblastoma - pathology; Neuroblastoma - radiotherapy; OCE; Optimization; Phantoms, Imaging; Planning; Predictive Value of Tests; Radiotherapy Dosage; Radiotherapy Planning, Computer-Assisted - methods; Radiotherapy, Intensity-Modulated - methods; Skull Neoplasms - radiotherapy; Skull Neoplasms - secondary; Technical Notes
• ISSN: 1526-9914; 1526-9914
• DOI: 10.1120/jacmp.v10i4.3061

IMRT plans generated in Eclipse use a fast algorithm to evaluate dose for optimization and a more accurate algorithm for a final dose calculation, the Analytical Anisotropic Algorithm. The use of a fast optimization algorithm introduces optimization convergence errors into an IMRT plan. Eclipse has a feature where optimization may be performed on top of an existing base plan. This feature allows for the possibility of arriving at a recursive solution to optimization that relies on the accuracy of the final dose calculation algorithm and not the optimizer algorithm. When an IMRT plan is used as a base plan for a second optimization, the second optimization can compensate for heterogeneity and modulator errors in the original base plan. Plans with the same field arrangement as the initial base plan may be added together by adding the initial plan optimal fluence to the dose correcting plan optimal fluence. A simple procedure to correct for optimization errors is presented that may be implemented in the Eclipse treatment planning system, along with an Excel spreadsheet to add optimized fluence maps together. PACS number: 87.53.Bn, 87.56.By