Author: Matthew Stephen Andriotty, Chi Ma, Xiao Wang, Keying (Karen) Xu, Suhong Yu, Ning J. Yue, Yin Zhang 👨🔬
Affiliation: Brigham and Women's Hospital, Rutgers Cancer Institute of New Jersey 🌍
Purpose: To evaluate the effects of overlapping CT reconstruction on dosimetry calculation for LINAC-based SRS patients simulated on CT simulators from three different vendors.
Methods: CT simulation scans for SRS patients were retrospectively reconstructed with varying slice thickness and interval combinations and compared with the non-overlapping CT. Treatment plans for each patient were copied onto each reconstructed CT image, and dose was re-calculated. To compare the effects of different reconstruction parameters, metrics such as PTV coverage and OAR doses were calculated for the plans on each CT image.
Results: Values of D0.035cc for various OARs were calculated for plans on each overlapping CT and compared with the values from the plan on the non-overlapping CT. The average difference of D0.035cc on overlapping CTs compared to that of the non-overlapping image was 2.2 cGy for the GE scanner, 15.7 cGy for the Siemens scanner, and 2.9 cGy for the Philips scanner.
Similarly, values of V100% for the PTVs were compared. V100% on the overlapping CTs differed from the non-overlapping image by an average of 0.6 percentage points for the GE scanner, 0.1 percentage points for the Siemens scanner, and 0.1 percentage points for the Philips scanner.
Conclusion: Overlapping reconstruction of CT simulation can reduce image noise by using thicker slices and increase z-direction resolution by using thinner reconstruction intervals. Dosimetry calculations for SRS patients, performed using three different CT simulators with overlapping reconstructions, showed only minor differences in key dosimetric parameters for both OARs and target structures. These findings support the use of overlapping reconstruction techniques to improve image quality while maintaining dosimetric accuracy.