Investigation of the Planning Strategy with Dual-Algorithm for Small Lesions in the Heterogeneous Region
M Lin*, J Li, L Wang, E Fourkal, R Price, J Fan, C Ma, Fox Chase Cancer Center, Philadelphia, PASU-E-T-705 Sunday 3:00PM - 6:00PM Room: Exhibit Hall
Purpose: This study performed an in-depth investigation of two SBRT planning strategies for small lung lesion treatment and evaluated their effectiveness of overcoming the heterogeneity impact on the plan quality. The first approach is to re-normalize the dose by prescribing to a different isodose level in order to achieve the desired target dose coverage after the initial plan is recalculated using the Monte Carlo (MC) method. The second approach is to re-optimize the beam weights of the initial plan based on the MC calculated dose.
Methods: Mutiplan that has both the MC calculation method and the ray-tracing algorithm was employed. Sixteen lung cancer patients (PTV4.7-58.5cm3) treated with Cyberknife were recruited. The ray-tracing algorithm was utilized for the initial plan optimization as the only choice provided by the Multiplan. Subsequently, the two planning strategies were performed followed by the dose distribution and plan quality comparison with the initial plan. The single beam dose calculations with the ray-tracing and the MC method were compared for investigating why re-optimization is able to improve the quality of the most plans but not all.
Results: Fourteen out of the sixteen cases demonstrate a better plan quality in the re-optimized plan. The average conformity index of the initial, the re-normalized, and the re-optimized plans are 1.14±0.08, 1.48±0.26, and 1.27±0.11, respectively, while the heterogeneity indexes are 1.28±0.07, 1.48±0.09, and 1.42±0.2. The comparison of the critical structure DVH parameters supports that the re-optimized plans can reduce the dose to the critical structures.
Conclusions: Both strategies provide plans fulfilling the clinical criteria. Re-optimization achieves a better plan quality, especially for irregular and peripherally located lesions. For small isolated- or low density lesions, re-normalization method provides better plan quality. Algorithms to better accounting for the heterogeneity effect are necessary for the initial plan optimization to further improve the plan quality.