Fractal Analysis for Assessing IMRT Modulation Complexity
Y Kakinohana*, T Toita, T Ariga, G Kasuya, School of Medicine, Ryukyu University OkinawaSU-E-T-698 Sunday 3:00PM - 6:00PM Room: Exhibit Hall
Purpose: To evaluate the feasibility of fractal dimension (FD as a modulation complexity metric in intensity modulated radiation therapy (IMRT) treatment fields.
Methods: Uniform sliding window (SW) deliveries changing the leaf gap width (1-50 mm) were utilized for this study. During the SW deliver, the dynamic treatment log file (Dynalog file) was acquired. Binary images were produced from the DynaLog files of the sliding window fields. For each sliding window field, five binary images changing the pixel size (0.1, 0.2, 0.3, 0.4 and 0.5 mm) were produced. The FD was evaluated by the Box-Counting method. The box sizes were 1, 2, 4, 8, 16 and 32 pixels. The fractal dimension for each segment and the overall field was evaluated by the averaging five images. In this study, FD is the average for all segments and the standard deviation of FD is the variation of FD between segments. The analysis was carried out by in-house programming in ImageJ (http://imagej.nih.gov/ij/ ). A highly modulated field may result in small leaf gaps, therefore low FD is expected for increased modulation complexity.
Results: The FD decreases with the decrease of leaf gap width. The FDs and SDs were 1.964±0.0007, 1.879±0.0054, 1.816±0.0092, 1.402±0.0429 for 50, 10, 5, and 1 mm gap, respectively.
The large SDs for small gap widths resulted from variability in leaf positions during the SW delivery. For 1 mm gap, a low FD value of 1.35 was observed at the middle segment where uneven leaf position could be observed visually.
Conclusion: The FD for smaller leaf gap which implies highly modulated was lower. This work indicates that the FD method could be used for assessing IMRT plan complexity. Further studies such as comparison to other metrics and its usefulness in clinical IMRT plans are needed.