Improved target visualization using a novel iterative CBCT reconstruction algorithm for prostate cancer RT.
Gardner S, Mao W, Liu C, Aref I, Elshaikh M, Lee J, Pradhan D, Siddiqui F, Movsas B, and Chetty I. Improved target visualization using a novel iterative CBCT reconstruction algorithm for prostate cancer RT. Med Phys 2018; 45(6):e485-e486.
Purpose: To use quantitative (contouring variability) and qualitative (image quality) means to evaluate the clinical utility of a novel iterative cone-beam CT (CBCT) reconstruction for imaging of prostate cancer patients. Methods: Nine prostate cancer patients were selected for this study. For each patient, iterative CBCT(iterative-CBCT) was compared to conventional CBCT(FDK-CBCT) using contouring and image quality analysis. For prostate contouring analysis, 4-5 experts contoured the prostate, seminal vesicles, bladder, and rectum on both image sets. Consensus contours were generated in CERR using the STAPLE method for use as reference standard. The following metrics were used to compare observer contours to consensus: Dice coefficient, Hausdorff distance, and Mean Contour Distance. The iterative- CBCT data was compared to FDK-CBCT data using t-test (P < 0.05 significant). For image quality analysis, 8 observers graded image sets on a scale ranging from: 1(indicating iterative-CBCT image quality is far superior to FDK-CBCT) to 5(indicating iterative-CBCT is far inferior to FDK-CBCT). Results: Contours on iterative-CBCT image displayed 4.4% improvement in Dice coefficient (P = 0.04) for prostate contour. Improvement in Mean Contour Distance was notable for the following: 1.3 mm average improvement in superior contour region (prostate-bladder interface), and 0.5 mm average improvement in posterior contour region (prostate-rectum interface). We also note a trend towards improvement of 1.5 mm in Hausdorff distance (P = 0.17). No statistically significant differences were found for seminal vesicles, bladder, or rectum contours. For prostate images, observers noted improved image quality for 53/72(73.6%) image evaluations and far superior image quality for 13/72(16.7%) image evaluations. Conclusion: Observers noted a marked improvement in image uniformity, noise level, and overall image quality for CBCT images generated using a novel iterative reconstruction algorithm. In addition, expert observers displayed improvement in the ability to consistently delineate the prostate. Thus, the novel iterative reconstruction algorithm analyzed in this study is capable of improving target visualization for prostate cancer RT.