Patients with arteriovenous malformation (n = 10) or acoustic neuromas (n = 5) were planned with different treatment modalities. Paddick conformity index (CI), dose heterogeneity (DH), gradient index (GI) and beam-on time were used as dosimetric indices.
Gamma-Knife-Perfexion can achieve high degree of conformity (CI = 0.77 ¡À 0.04) with limited low-doses (GI = 2.59 ¡À 0.10) surrounding the inhomogeneous dose distribution (DH = 0.84 ¡À 0.05) at the cost of treatment time (68.1 min ¡À 27.5). Novalis-Tx-DCA improved this inhomogeneity (DH = 0.30 ¡À 0.03) and treatment time (16.8 min ¡À 2.2) at the cost of conformity (CI = 0.66 ¡À 0.04) and Novalis-TX-DMLC-IMRT improved the DCA CI (CI = 0.68 ¡À 0.04) and inhomogeneity (DH = 0.18 ¡À 0.05) at the cost of low-doses (GI = 3.94 ¡À 0.92) and treatment time (21.7 min ¡À 3.4) (p < 0.01). Cyberknife achieved comparable conformity (CI = 0.77 ¡À 0.06) at the cost of low-doses (GI = 3.48 ¡À 0.47) surrounding the homogeneous (DH = 0.22 ¡À 0.02) dose distribution and treatment time (28.4 min ¡À 8.1) (p < 0.01).
Gamma-Knife-Perfexion will comply with all SRS constraints (high conformity while minimizing low-dose spread). Multiple focal entries (Gamma-Knife-Perfexion and Cyberknife) will achieve better conformity than High-Definition-MLC of Novalis-Tx at the cost of treatment time. Non-isocentric beams (Cyberknife) or IMRT-beams (Novalis-Tx-DMLC-IMRT) will spread more low-dose than multiple isocenters (Gamma-Knife-Perfexion) or dynamic arcs (Novalis-Tx-DCA). Inverse planning and modulated fluences (Novalis-Tx-DMLC-IMRT and CyberKnife) will deliver the most homogeneous treatment. Furthermore, Linac-based systems (Novalis and Cyberknife) can perform image verification at the time of treatment delivery.