AFI manual planning versus HyperArc auto‐planning: A head‐to‐head comparison of SRS plan quality

• Published in: Journal of applied clinical medical physics Vol. 25; no. 11; pp. e14503 - n/a
• Main Authors: Desai, Dharmin D., Cordrey, Ivan L., Johnson, E. Lee, Oldland, Thomas A.
• Format: Journal Article
• Language: English
• Published: United States John Wiley & Sons, Inc 01.11.2024; John Wiley and Sons Inc
• Subjects: AFI; Automation; Fractionation; Fractions; HyperArc; Optimization; plan quality; Planning; R50; Radiation Oncology Physics; Radiation therapy; Radiosurgery; SIMT; Software; SRS; Palo Alto California; United States > US; SIMT; AFI; R50; SRS; plan quality; HyperArc
• ISSN: 1526-9914; 1526-9914
• DOI: 10.1002/acm2.14503

Introduction HyperArc (HA) auto‐planning offers simplicity for the end user and consistently high‐quality SRS plans. The “Ask For It” (AFI) optimization strategy offers a manual planning technique that, when coupled with R50%Analytic, can be guided to deliver a plan with an intermediate dose spill “as low as reasonably achievable” and high target dose conformity. A direct comparison of SRS plan quality obtained using the manual planning AFI strategy and HA has been performed. Methods Using a CT data set available from the Radiosurgery Society, 54 PTVs were created and used to generate 19 individual SRS/SRT cases. Case complexity ranged from single PTV plans to multiple PTV plans with a single isocenter. PTV locations ranged from relative isolation from critical structures to lesions within 1.5 mm of the optic apparatus and abutting the brainstem. All cases were planned using both the AFI and HA optimization strategies as implemented in the Varian Medical Systems Eclipse Treatment Planning System. A range of treatment plan quality metrics were obtained including Intermediate Dose Spill (R50%), Conformity Indices CIRTOG and CIPaddick, PTV Dose Coverage (Dn%), PTV Mean Dose, and Modulation Factor. The Wilcoxon Signed Rank Sum non‐parametric statistical method was utilized to compare the obtained plan quality metrics. Results Statistically significant improvements were found for the AFI strategy for metrics R50%, CIRTOG, CIPaddick, and PTV Mean Dose (p < 0.001). HA achieved superior coverage for Dn% (p = 0.018), while the Modulation Factors were not significantly different for AFI compared to HA optimization (p = 0.13). Conclusion This study provides evidence that the AFI manual planning strategy can produce high‐quality planning metrics similar to the HA auto‐planning method.