On the role of the optimization algorithm of RapidArc® volumetric modulated arc therapy on plan quality and efficiency

• Published in: Medical physics (Lancaster) Vol. 38; no. 11; pp. 5844 - 5856
• Main Authors: Vanetti, Eugenio, Nicolini, Giorgia, Nord, Janne, Peltola, Jarkko, Clivio, Alessandro, Fogliata, Antonella, Cozzi, Luca
• Format: Journal Article
• Language: English
• Published: United States American Association of Physicists in Medicine 01.11.2011
• Subjects: Algorithms; Anatomy; Brain; Cancer; Dose‐volume analysis; dosimetry; Humans; Information and communication theory; inverse problems; Iteration theory; Lungs; medical computing; Multileaf collimators; Neoplasms - pathology; Neoplasms - radiotherapy; optimisation; Optimization; Photons; progressive resolution algorithm; Quality assurance; Quality Control; radiation therapy; Radiotherapy Dosage; Radiotherapy Planning, Computer-Assisted - methods; Radiotherapy Planning, Computer-Assisted - standards; RapidArc; Therapeutics; tumours; volumetric modulated arc therapy; progressive resolution algorithm; volumetric modulated arc therapy; RapidArc
• ISSN: 0094-2405; 2473-4209
• DOI: 10.1118/1.3641866

Purpose: The RapidArc volumetric modulated arc therapy (VMAT) planning process is based on a core engine, the so-called progressive resolution optimizer (PRO). This is the optimization algorithm used to determine the combination of field shapes, segment weights (with dose rate and gantry speed variations), which best approximate the desired dose distribution in the inverse planning problem. A study was performed to assess the behavior of two versions of PRO. These two versions mostly differ in the way continuous variables describing the modulated arc are sampled into discrete control points, in the planning efficiency and in the presence of some new features. The analysis aimed to assess (i) plan quality, (ii) technical delivery aspects, (iii) agreement between delivery and calculations, and (iv) planning efficiency of the two versions. Methods: RapidArc plans were generated for four groups of patients (five patients each): anal canal, advanced lung, head and neck, and multiple brain metastases and were designed to test different levels of planning complexity and anatomical features. Plans from optimization with PRO2 (first generation of RapidArc optimizer) were compared against PRO3 (second generation of the algorithm). Additional plans were optimized with PRO3 using new features: the jaw tracking, the intermediate dose and the air cavity correction options. Results: Results showed that (i) plan quality was generally improved with PRO3 and, although not for all parameters, some of the scored indices showed a macroscopic improvement with PRO3. (ii) PRO3 optimization leads to simpler patterns of the dynamic parameters particularly for dose rate. (iii) No differences were observed between the two algorithms in terms of pretreatment quality assurance measurements and (iv) PRO3 optimization was generally faster, with a time reduction of a factor approximately 3.5 with respect to PRO2. Conclusions: These results indicate that PRO3 is either clinically beneficial or neutral in terms of dosimetric quality while it showed significant advantages in speed and technical aspects.