187. Evaluation of intrarater and interrater VMAT plans obtained with MCO algorithm and comparison with DMPO plans. (December 2018)

• Record Type: Journal Article
• Authors: de Fatis, P. Tabarelli; Liotta, M.; Malovini, A.; Ugolini, M.
• Is Part Of: Physica medica. Volume 56(2018)Supplement 2
• Page Start: 176
• Page End: 177
• Publication Date: 2018-12
• Subjects: 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.2018.04.198
• Languages: English
• ISSNs: 1120-1797

Abstract

Abstract: Purpose: to evaluate Multicriteria Optimization Algorithm (MCO) and its reliability in providing VMAT treatment plans through the technique of Pareto surfaces navigation. Methods: ten patients who were previously treated with VMAT for prostate cancer were selected and replanned five times each by three different operators (total 50 plans) in RayStation treatment planning system using MCO[1] that provides a real-time assessment of the tradeoff between different clinical goals. For the optimization of each plan we started from the same template derived according the target coverage and the spare of healthy tissue used in our clinical protocols. Furthermore we calculate also a manual plan for each patient using the algorithm based on Direct Machine Parameter Optimization (DMPO); the quality of the plans was evaluated through D2%, D98%, Conformity Index (CI = [volume that receives 95% of dose prescription/target volume]) and Homogeneity index (HI = [dose that covers 5% of the target/dose that covers 95% of the target]). We evaluated intrarater and interrater reliability of the MCO plans through Intraclass Correlation Coefficient (ICC). Finally we compared the best MCO plan (assigning a score to each of the four indexes) with the one obtained with DMPO optimization estimating Pearson and Linn's Correlation Coefficients (PCC and LCC). Results: The average intrarater ICC over the three operators for CI index was "fair" (0, 50). Interrater ICC showed "poor" similarity between operators for CI (0, 35). Both PCC and LCC between DMPO plan and MCO best plan were relevant (⩾0, 7) for CI for op3. Conclusions: Despite some "poor" statistical similarity found in the analysis, the MCO optimization has the potential to provide high-quality plans comparable to the DMPO algorithm in terms of PTV coverage especially in isodoses conformity achievement.