Room: Karl Dean Ballroom A1
Purpose: To develop a treatment plan quality control methodology that utilizes a large set of previously treated radiotherapy plans. Past patient DVH and anatomical data are statistically analyzed to account for variations in patientsâ€™ anatomies.
Methods: Interpatient anatomical variations have restricted direct comparisons of multiple plans. We analyze five sources of such variations. For the tumor, we compute surface area, volume, and spread (maximum distance in x, y, z dimensions). Additionally, we compute minimum distance, and center of mass distance, between tumor and organ. Organ DVH points Dâ‚“ are used as proxies for plan quality. Linear regression is used to model the effect of the sources on Dâ‚“. The residuals from the regression model are used to construct anatomy-adjusted individual control charts (I-Charts). These charts identify patients with out-of-control DVH points (more than three standard deviations away from the mean) after accounting for individual variations in anatomies. Sixty-nine head-and-neck cases are used for the evaluation of the proposed method and the results are compared to the conventional I-Charts.
Results: For all patients, the brainstem and spinal cord were the primary organs to spare. For the brainstem, the conventional I-Chart identified patients 4, 6, 43, and 58 with 30.9Gy, 28.9Gy, 18.6Gy and 23.5Gy for Dâ‚‰â‚ˆ as out-of-control. However, the anatomy-adjusted I-Chart identified only patients 4 and 6. Patients 43 and 58 were no longer considered out-of-control once the center of mass distance between the tumor and organ (most significant contributor towards brainstem Dâ‚‰â‚ˆ) was accounted for in the adjusted I-Chart. Similar results are observed for spinal cord Dâ‚‚.
Conclusion: An adjusted I-Chart is developed for monitoring plan quality that is robust to interpatient variations in anatomical geometry. This framework is generic and can readily be incorporated to account for other sources of interpatient variations.
Quality Control, Dose Volume Histograms, Statistical Analysis