SU‐E‐T‐867: Improving the IMRT Plan Quality by Minimizing the Discrepancy Between the Optimized and Segmented Intensity Maps

Abstract

Purpose: In step‐and‐shoot IMRT, beam intensity maps are first optimized, followed by a leaf‐sequencing to translate them into deliverable segments. The differences between the optimized and segmented intensity maps may degrade the deliverable dose distribution. Aiming to decrease this degradation, we present a method to minimize the discrepancy between the optimized and segmented intensity maps. Methods: The general idea of our algorithm is to minimize the sum of absolute differences (SOAD) between the optimized and sequenced intensity maps, similar to the work of Daliang Cao et al (Med. Phys. 2006, 33(4): 859–867). The improvement of our method includes: (1) the initial segment shapes and weights are produced based on the optimized intensity maps with a fast Xia and Verheyˈs leaf sequencing algorithm; (2) the segment weights and shapes are iteratively optimized with a modified gradient conjugate algorithm and genetic algorithm; and (3) the MLC‐dependent leaf transmission and penumbra are both accounted for. Results: For a simulated case involving the left and right lungs and an organ at risk (OAR) surrounded by a C‐shaped planning target volume (PTV), five beams were used. For the five optimized beam intensity maps, the segmented maps were computed using the proposed algorithm. The average computation time was about 17 sec per beam. Both of the dose distributions and dose volume histograms (DVHs) showed the improvements given by our method: more uniform dose distributions for PTV of the both cases, and lower doses for OARs of the clinical case, than that when the MLC‐dependent leaf transmission and penumbra were ignored. Also, the needed numbers of segment and monitor unit were reduced. Conclusions: The proposed method could produce a better segmentation result by improving the dose distributions and reducing the segment number.Funding support: This work was supported by the Program for New Century Excellent Talents in University of China (NCET‐07‐0151) and the Natural Science Foundations of China (61075109, 30730036).