Tiago Ventura1, Humberto Rocha2, Brigida da Costa Ferreira3, Joana Dias4, Maria do Carmo Lopes5. 1. Physics Department of University of Aveiro, Campus Universitário de Santiago, 3810-193 Aveiro, Portugal; Medical Physics Department of Instituto Português de Oncologia de Coimbra Francisco Gentil, EPE, Avenida Bissaya Barreto, n° 98, 3000-075 Coimbra, Portugal; Institute for Systems Engineering and Computers at Coimbra, Rua Sílvio Lima, 3030-290 Coimbra, Portugal. Electronic address: tiagoventura@ipocoimbra.min-saude.pt. 2. Institute for Systems Engineering and Computers at Coimbra, Rua Sílvio Lima, 3030-290 Coimbra, Portugal; Economy Faculty of University of Coimbra and Centre for Business and Economics Research, Avenida Dr. Dias da Silva, nº 165, 3004-512 Coimbra, Portugal. Electronic address: hrocha@mat.uc.pt. 3. Institute for Systems Engineering and Computers at Coimbra, Rua Sílvio Lima, 3030-290 Coimbra, Portugal; School Health Polytechnic of Porto, Rua Dr. António Bernardino de Almeida, nº 400, 4200-072 Porto, Portugal; I3N Physics Department of University of Aveiro, Campus Universitário de Santiago, 3810-193 Aveiro, Portugal. Electronic address: bcf@ess.ipp.pt. 4. Institute for Systems Engineering and Computers at Coimbra, Rua Sílvio Lima, 3030-290 Coimbra, Portugal; Economy Faculty of University of Coimbra and Centre for Business and Economics Research, Avenida Dr. Dias da Silva, nº 165, 3004-512 Coimbra, Portugal. Electronic address: joana@fe.uc.pt. 5. Physics Department of University of Aveiro, Campus Universitário de Santiago, 3810-193 Aveiro, Portugal; Medical Physics Department of Instituto Português de Oncologia de Coimbra Francisco Gentil, EPE, Avenida Bissaya Barreto, n° 98, 3000-075 Coimbra, Portugal; Institute for Systems Engineering and Computers at Coimbra, Rua Sílvio Lima, 3030-290 Coimbra, Portugal. Electronic address: mclopes@ipocoimbra.min-saude.pt.
Abstract
PURPOSE: To compare two beam angle optimization (BAO) algorithms for coplanar and non-coplanar geometries in a multicriterial optimization framework. METHODS: 40 nasopharynx patients were selected for this retrospective planning study. IMRT optimized plans were produced by Erasmus-iCycle multicriterial optimization platform. Two different algorithms, based on a discrete and on a continuous exploration of the space search, algorithm i and B respectively, were used to address BAO. Plan quality evaluation and comparison were performed with SPIDERplan. Statistically significant differences between the plans were also assessed. RESULTS: For plans using only coplanar incidences, the optimized beam distribution with algorithm i is more asymmetric than with algorithm B. For non-coplanar beam optimization, larger deviations from coplanarity were obtained with algorithm i than with algorithm B. Globally, both algorithms presented near equivalent plan quality scores, with algorithm B presenting a marginally better performance than algorithm i. CONCLUSION: Almost all plans presented high quality, profiting from multicriterial and beam angular optimization. Although there were not significant differences when average results over the entire sample were considered, a case-by-case analysis revealed important differences for some patients.
PURPOSE: To compare two beam angle optimization (BAO) algorithms for coplanar and non-coplanar geometries in a multicriterial optimization framework. METHODS: 40 nasopharynx patients were selected for this retrospective planning study. IMRT optimized plans were produced by Erasmus-iCycle multicriterial optimization platform. Two different algorithms, based on a discrete and on a continuous exploration of the space search, algorithm i and B respectively, were used to address BAO. Plan quality evaluation and comparison were performed with SPIDERplan. Statistically significant differences between the plans were also assessed. RESULTS: For plans using only coplanar incidences, the optimized beam distribution with algorithm i is more asymmetric than with algorithm B. For non-coplanar beam optimization, larger deviations from coplanarity were obtained with algorithm i than with algorithm B. Globally, both algorithms presented near equivalent plan quality scores, with algorithm B presenting a marginally better performance than algorithm i. CONCLUSION: Almost all plans presented high quality, profiting from multicriterial and beam angular optimization. Although there were not significant differences when average results over the entire sample were considered, a case-by-case analysis revealed important differences for some patients.