An effective method for smoothing the staggered dose distribution of multi-leaf collimator field edge

Purpose: To smooth the staggered dose distribution that occurs in stepped leaves defined by a multi-leaf collimator (MLC). als and methods: The MLC Shaper program controlled the stepped leaves, which were shifted in a traveling range, the pattern of shift was from the position of out-bound to in-bound with a one-segment (cross-bound), three-segment, and five-segment shifts. Film was placed at a depth of 1.5 cm and irradiated with the same irradiation dose used for the cerrobend block experiment. Four field edges with the MLC defining at 15°, 30°, 45°, 60° angels relative to the jaw edge were performed, respectively, in this study. For the field edge defined by the multi-segment technique, the amplitude of the isodose lines for 50% isodose line and both the 80% and 20% isodose lines were measured. The effective penumbra widths with 90–10% and 80–20% distances for different irradiations were determined at four field edges with the MLC defining at 15°, 30°, 45°, 60° angels relative to the jaw edge. s: Use of the five-segment technique for multi-leaf collimation at the 60° angle field edge smoothes each isodose line into an effectively straight line, similar to the pattern achieved using a cerrobend block. The separation of these lines is also important. The 80–20% effective penumbra width with five-segment techniques (8.23 mm) at 60° angle relative to the jaw edge is little wider (1.9 times) than the penumbra of cerrobend block field edge (4.23 mm). We also found that the 90–10% effective penumbra width with five-segment techniques (12.68 mm) at 60° angle relative to the jaw edge is little wider (1.28 times) than the penumbra of cerrobend block field edge (9.89 mm). sion: The multi-segment technique is effective in smoothing the MLC staggered field edge. The effective penumbra width with more segment techniques at larger degree angles relative to the field edge is little wider than the penumbra for a cerrobend block.